I was trying to keep track of the better post-leak code-analysis links on exactly this question, so I collected them here:\nhttps://github.com/nblintao/awesome-claude-code-postleak-ins...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:09:19.000Z","created_at_i":1774948159,"id":47584622,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Are there any interesting/uniq features present in it that are not in the alternatives? My understanding is that its just a client for the powerful llm","title":null,"type":"comment","url":null},{"author":"LeoDaVibeci","children":[{"author":"yellow_lead","children":[],"created_at":"2026-03-31T09:14:09.000Z","created_at_i":1774948449,"id":47584659,"options":[],"parent_id":47584640,"points":null,"story_id":47584540,"text":"No","title":null,"type":"comment","url":null},{"author":"avaer","children":[],"created_at":"2026-03-31T09:14:22.000Z","created_at_i":1774948462,"id":47584664,"options":[],"parent_id":47584640,"points":null,"story_id":47584540,"text":"No, it was never open source. You could always reverse engineer the cli app but you didn't have access to the source.","title":null,"type":"comment","url":null},{"author":"agluszak","children":[],"created_at":"2026-03-31T09:15:15.000Z","created_at_i":1774948515,"id":47584674,"options":[],"parent_id":47584640,"points":null,"story_id":47584540,"text":"You may have mistaken it with Codex
https://github.com/openai/codex","title":null,"type":"comment","url":null},{"author":"dragonwriter","children":[{"author":"alkonaut","children":[],"created_at":"2026-03-31T15:08:50.000Z","created_at_i":1774969730,"id":47588484,"options":[],"parent_id":47584710,"points":null,"story_id":47584540,"text":"> its not even source available
It _wasn't_ even source available.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:20:03.000Z","created_at_i":1774948803,"id":47584710,"options":[],"parent_id":47584640,"points":null,"story_id":47584540,"text":"> Isn't it open source?
No, its not even source available,.
> Or is there an open source front-end and a closed backend?
No, its all proprietary. None of it is open source.","title":null,"type":"comment","url":null},{"author":"karimf","children":[{"author":"matheusmoreira","children":[],"created_at":"2026-03-31T10:00:33.000Z","created_at_i":1774951233,"id":47584996,"options":[],"parent_id":47584887,"points":null,"story_id":47584540,"text":"Wow it's true. Anthropic actually had me fooled. I saw the GitHub repository and just assumed it was open source. Didn't look at the actual files too closely. There's pretty much nothing there.
So glad I took the time to firejail this thing before running it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:46:44.000Z","created_at_i":1774950404,"id":47584887,"options":[],"parent_id":47584640,"points":null,"story_id":47584540,"text":"The Github repo is only for issue tracker","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:11:59.000Z","created_at_i":1774948319,"id":47584640,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Isn't it open source?
Or is there an open source front-end and a closed backend?","title":null,"type":"comment","url":null},{"author":"treexs","children":[{"author":"avaer","children":[{"author":"JohnLocke4","children":[],"created_at":"2026-03-31T10:27:50.000Z","created_at_i":1774952870,"id":47585233,"options":[],"parent_id":47584953,"points":null,"story_id":47584540,"text":"You heard it here first","title":null,"type":"comment","url":null},{"author":"ares623","children":[],"created_at":"2026-03-31T10:32:10.000Z","created_at_i":1774953130,"id":47585276,"options":[],"parent_id":47584953,"points":null,"story_id":47584540,"text":"So close to April Fool's too. I'm sure it will still be a surprise for a majority of their users.","title":null,"type":"comment","url":null},{"author":"cmontella","children":[{"author":"hansonkd","children":[{"author":"cmontella","children":[],"created_at":"2026-03-31T20:17:54.000Z","created_at_i":1774988274,"id":47592892,"options":[],"parent_id":47590268,"points":null,"story_id":47584540,"text":"If it could somehow only work if I maintain the kitchen sink and counter, then maybe I'd be motivated to keep the house clean. The gacha game trains you.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:58:09.000Z","created_at_i":1774976289,"id":47590268,"options":[],"parent_id":47586122,"points":null,"story_id":47584540,"text":"You know, that would actually be pretty fun and cool. Like if you had home automation set up with a "pet assistant", but it would only follow your commands if you made sure to keep it happy.","title":null,"type":"comment","url":null},{"author":"financetechbro","children":[{"author":"cmontella","children":[],"created_at":"2026-03-31T20:16:45.000Z","created_at_i":1774988205,"id":47592877,"options":[],"parent_id":47592432,"points":null,"story_id":47584540,"text":"haha right! I think if I were to make Rocky though, there are a number of more geometric symbols. Maybe like
\u256d\u2b1f\u256eThe joke was the assistant is a cat who is constantly sabotaging you, and you have to take care of it like a gacha pet.
The seriousness though is that actually, disembodied intelligences are weird, so giving them a face and a body and emotions is a natural thing, and we already see that with various AI mascots and characters coming into existence.
[1]: serious: https://github.com/mech-lang/mech/releases/tag/v0.3.1-beta
[2]: joke: https://github.com/cmontella/purrtran","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:55:19.000Z","created_at_i":1774950919,"id":47584953,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"(spoiler alert)
Buddy system is this year's April Fool's joke, you roll your own gacha pet that you get to keep. There are legendary pulls.
They expect it to go viral on Twitter so they are staggering the reveals.","title":null,"type":"comment","url":null},{"author":"TIPSIO","children":[{"author":"redrove","children":[],"created_at":"2026-03-31T13:37:28.000Z","created_at_i":1774964248,"id":47587201,"options":[],"parent_id":47586379,"points":null,"story_id":47584540,"text":"Not necessarily; I would very much like to use those features on a Linux server. Currently the Anthropic implementation forces a desktop (or worse, a laptop) to be turned on instead of working headless as far as I understand it.
I\u2019ll give clappie a go, love the theme for the landing page!","title":null,"type":"comment","url":null},{"author":"barbazoo","children":[{"author":"TIPSIO","children":[],"created_at":"2026-03-31T13:55:48.000Z","created_at_i":1774965348,"id":47587455,"options":[],"parent_id":47587330,"points":null,"story_id":47584540,"text":"Not at all. I am a big a Claude Code fan and glad they are releasing more and more features for users","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:45:48.000Z","created_at_i":1774964748,"id":47587330,"options":[],"parent_id":47586379,"points":null,"story_id":47584540,"text":"Poor mum","title":null,"type":"comment","url":null},{"author":"sanex","children":[{"author":"TIPSIO","children":[],"created_at":"2026-03-31T15:59:35.000Z","created_at_i":1774972775,"id":47589350,"options":[],"parent_id":47588822,"points":null,"story_id":47584540,"text":"Thanks so much! It's a fancy landing page thanks to Claude.
Tmux is seriously an amazing tool.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:29:20.000Z","created_at_i":1774970960,"id":47588822,"options":[],"parent_id":47586379,"points":null,"story_id":47584540,"text":"Clappie looks much more fabulous than CC though. I'll have to give it a try. I like how you put the requests straight into an already running CC session instead of calling `claude -p` every time like the claws.","title":null,"type":"comment","url":null},{"author":"Narretz","children":[],"created_at":"2026-03-31T17:48:03.000Z","created_at_i":1774979283,"id":47591019,"options":[],"parent_id":47586379,"points":null,"story_id":47584540,"text":"Dispatch and scheduled tasks have been available for a few weeks already, although with limitations.","title":null,"type":"comment","url":null},{"author":"theintern","children":[],"created_at":"2026-03-31T21:42:09.000Z","created_at_i":1774993329,"id":47593851,"options":[],"parent_id":47586379,"points":null,"story_id":47584540,"text":"I didn't know this existed. I had claude put this together for me from scratch a week ago. Tmux and claude and telegram is a really powerful combo!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:30:39.000Z","created_at_i":1774960239,"id":47586379,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"If this true. My old personal agent Claude Code setup I open sourced last month will finally be obsolete (1 month lol):
- Telegram Integration => CC Dispatch
- Crons => CC Tasks
- Animated ASCII Dog => CC Buddy","title":null,"type":"comment","url":null},{"author":"denimnerd42","children":[],"created_at":"2026-03-31T12:46:12.000Z","created_at_i":1774961172,"id":47586572,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"all these flags are findable by pointing claude at the binary and asking it to find festure flags.","title":null,"type":"comment","url":null},{"author":"charcircuit","children":[],"created_at":"2026-03-31T12:50:35.000Z","created_at_i":1774961435,"id":47586617,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"People already can look at the source without this leak. People have had hacked builds force enabling feature flags for a long time.","title":null,"type":"comment","url":null},{"author":"mghackerlady","children":[],"created_at":"2026-03-31T13:10:56.000Z","created_at_i":1774962656,"id":47586859,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"one of those is adorable and the other one is unethical","title":null,"type":"comment","url":null},{"author":"t0mas88","children":[{"author":"20k","children":[],"created_at":"2026-03-31T14:23:42.000Z","created_at_i":1774967022,"id":47587840,"options":[],"parent_id":47587498,"points":null,"story_id":47584540,"text":"This seems like a good way to weed out models: ask them to include the term capybara in their commit messages","title":null,"type":"comment","url":null},{"author":"jasonlotito","children":[],"created_at":"2026-03-31T16:13:26.000Z","created_at_i":1774973606,"id":47589584,"options":[],"parent_id":47587498,"points":null,"story_id":47584540,"text":"At least this was known with the Mythos "early blog post" fiasco.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:58:07.000Z","created_at_i":1774965487,"id":47587498,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"Note also the "Claude Capybara" reference in the undercover prompt: https://github.com/chatgptprojects/claude-code/blob/642c7f94...","title":null,"type":"comment","url":null},{"author":"baxtr","children":[],"created_at":"2026-03-31T18:01:49.000Z","created_at_i":1774980109,"id":47591187,"options":[],"parent_id":47584683,"points":null,"story_id":47584540,"text":"Is there an AGI mode FF? Asking for a friend\u2026","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:17:10.000Z","created_at_i":1774948630,"id":47584683,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"The big loss for Anthropic here is how it reveals their product roadmap via feature flags. A big one is their unreleased "assistant mode" with code name kairos.
Just point your agent at this codebase and ask it to find things and you'll find a whole treasure trove of info.
Edit: some other interesting unreleased/hidden features
- The Buddy System: Tamagotchi-style companion creature system with ASCII art sprites
- Undercover mode: Strips ALL Anthropic internal info from commits/PRs for employees on open source contributions","title":null,"type":"comment","url":null},{"author":"q3k","children":[{"author":"loevborg","children":[{"author":"Insensitivity","children":[{"author":"luc_","children":[{"author":"werdnapk","children":[],"created_at":"2026-03-31T11:42:30.000Z","created_at_i":1774957350,"id":47585901,"options":[],"parent_id":47584817,"points":null,"story_id":47584540,"text":"insert "AI is just if statements" meme","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:34:20.000Z","created_at_i":1774949660,"id":47584817,"options":[],"parent_id":47584780,"points":null,"story_id":47584540,"text":"Fits with the origin story of Claude Code...","title":null,"type":"comment","url":null},{"author":"loevborg","children":[{"author":"Insensitivity","children":[],"created_at":"2026-03-31T09:50:33.000Z","created_at_i":1774950633,"id":47584924,"options":[],"parent_id":47584899,"points":null,"story_id":47584540,"text":"Maybe, I do suspect _some_ parts are codegen or source map artifacts.
But if you take a look at the other file, for example `useTypeahead` you'd see, even if there are a few code-gen / source-map artifacts, you still see the core logic, and behavior, is just a big bowl of soup","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:47:59.000Z","created_at_i":1774950479,"id":47584899,"options":[],"parent_id":47584780,"points":null,"story_id":47584540,"text":"useCanUseTool.tsx looks special, maybe it'scodegen'ed or copy 'n pasted? `_c` as an import name, no comments, use of promises instead of async function. Or maybe it's just bad vibing...","title":null,"type":"comment","url":null},{"author":"Overpower0416","children":[{"author":"worksonmine","children":[{"author":"Overpower0416","children":[],"created_at":"2026-03-31T11:31:11.000Z","created_at_i":1774956671,"id":47585811,"options":[],"parent_id":47585641,"points":null,"story_id":47584540,"text":"Using guard clauses. Way more readable and easy to work with.
export function extractSearchToken(completionToken: {\n token: string;\n isQuoted?: boolean;\n }): string {\n if (completionToken.isQuoted) {\n return completionToken.token.slice(2).replace(/"$/, '');\n }\n if (completionToken.token.startsWith('@')) {\n return completionToken.token.substring(1);\n }\n return completionToken.token;\n }What is the problem with that? How would you write that snippet? It is common in the new functional js landscape, even if it is pass-by-ref.","title":null,"type":"comment","url":null},{"author":"kelnos","children":[{"author":"catlifeonmars","children":[],"created_at":"2026-03-31T11:48:38.000Z","created_at_i":1774957718,"id":47585970,"options":[],"parent_id":47585832,"points":null,"story_id":47584540,"text":"I\u2019m not strongly opinionated, especially with such a short function, but in general early return makes it so you don\u2019t need to keep the whole function body in your head to understand the logic. Often it saves you having to read the whole function body too.
But you can achieve a similar effect by keeping your functions small, in which case I think both styles are roughly equivalent.","title":null,"type":"comment","url":null},{"author":"whilenot-dev","children":[],"created_at":"2026-03-31T12:00:51.000Z","created_at_i":1774958451,"id":47586086,"options":[],"parent_id":47585832,"points":null,"story_id":47584540,"text":"> This approximates `if` statements being an expression that returns something.
Do you care to elaborate? "if (...) return ...;" looks closer to an expression for me:
export function extractSearchToken(completionToken: { token: string; isQuoted?: boolean }): string {\n if (completionToken.isQuoted) return completionToken.token.slice(2).replace(/"$/, '');\n\n if (completionToken.token.startsWith('@')) return completionToken.token.substring(1);\n\n return completionToken.token;\n } export function extractSearchToken(completionToken: {\n token: string;\n isQuoted?: boolean;\n }): string {\n if (completionToken.isQuoted) {\n // Remove @" prefix and optional closing "\n return completionToken.token.slice(2).replace(/"$/, '');\n } else if (completionToken.token.startsWith('@')) {\n return completionToken.token.substring(1);\n } else {\n return completionToken.token;\n }\n }\nIt's extremely nested, it's basically an if statement soup
`useTypeahead.tsx` is even worse, extremely nested, a ton of "if else" statements, I doubt you'd look at it and think this is sane code","title":null,"type":"comment","url":null},{"author":"q3k","children":[{"author":"loevborg","children":[{"author":"darkstar_16","children":[],"created_at":"2026-03-31T10:13:47.000Z","created_at_i":1774952027,"id":47585125,"options":[],"parent_id":47584922,"points":null,"story_id":47584540,"text":"They use claude code to code it. Makes sense","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:50:32.000Z","created_at_i":1774950632,"id":47584922,"options":[],"parent_id":47584832,"points":null,"story_id":47584540,"text":"You're right about process.argv - wow, that looks like a maintenance and testability nightmare.","title":null,"type":"comment","url":null},{"author":"s3p","children":[],"created_at":"2026-03-31T10:05:55.000Z","created_at_i":1774951555,"id":47585051,"options":[],"parent_id":47584832,"points":null,"story_id":47584540,"text":"It probably exists only in CLAUDE or AGENTS.md since no humans are working on the code!","title":null,"type":"comment","url":null},{"author":"delamon","children":[{"author":"hu3","children":[],"created_at":"2026-03-31T10:11:33.000Z","created_at_i":1774951893,"id":47585101,"options":[],"parent_id":47585061,"points":null,"story_id":47584540,"text":"For one it's harder to unit test.","title":null,"type":"comment","url":null},{"author":"q3k","children":[],"created_at":"2026-03-31T10:52:19.000Z","created_at_i":1774954339,"id":47585453,"options":[],"parent_id":47585061,"points":null,"story_id":47584540,"text":"It's implicit state that's also untyped - it's just a String -> String map without any canonical single source of truth about what environment variables are consulted, when, why and in what form.
Such state should be strongly typed, have a canonical source of truth (which can then be also reused to document environment variables that the code supports, and eg. allow reading the same options from configs, flags, etc) and then explicitly passed to the functions that need it, eg. as function arguments or members of an associated instance.
This makes it easier to reason about the code (the caller will know that some module changes its functionality based on some state variable). It also makes it easier to test (both from the mechanical point of view of having to set environment variables which is gnarly, and from the point of view of once again knowing that the code changes its behaviour based on some state/option and both cases should probably be tested).","title":null,"type":"comment","url":null},{"author":"withinboredom","children":[{"author":"efskap","children":[{"author":"withinboredom","children":[],"created_at":"2026-04-02T06:08:38.000Z","created_at_i":1775110118,"id":47610544,"options":[],"parent_id":47610249,"points":null,"story_id":47584540,"text":"The process itself (including other threads) can call setenv whenever it wants.","title":null,"type":"comment","url":null}],"created_at":"2026-04-02T05:17:06.000Z","created_at_i":1775107026,"id":47610249,"options":[],"parent_id":47585797,"points":null,"story_id":47584540,"text":"Wait, is it expected for them to be able to change? According to this SO answer [0] it's only really possible through GDB or "nasty hacks" as there's no API for it.
[0] https://unix.stackexchange.com/questions/38205/change-enviro...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:28:27.000Z","created_at_i":1774956507,"id":47585797,"options":[],"parent_id":47585061,"points":null,"story_id":47584540,"text":"environment variables can change while the process is running and are not memory safe (though I suspect node tries to wrap it with a lock). Meaning if you check a variable at point A, enter a branch and check it again at point B ... it's not guaranteed that they will be the same value. This can cause you to enter "impossible conditions".","title":null,"type":"comment","url":null},{"author":"lioeters","children":[],"created_at":"2026-03-31T11:30:26.000Z","created_at_i":1774956626,"id":47585806,"options":[],"parent_id":47585061,"points":null,"story_id":47584540,"text":"> process.env? It is a global map
That's exactly why, access to global mutable state should be limited to as small a surface area as possible, so 99% of code can be locally deterministic and side-effect free, only using values that are passed into it. That makes testing easier too.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:06:55.000Z","created_at_i":1774951615,"id":47585061,"options":[],"parent_id":47584832,"points":null,"story_id":47584540,"text":"What is wrong with peeking at process.env? It is a global map, after all. I assume, of course, that they don't mutate it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:36:42.000Z","created_at_i":1774949802,"id":47584832,"options":[],"parent_id":47584743,"points":null,"story_id":47584540,"text":"
1. Randomly peeking at process.argv and process.env all around. Other weird layering violations, too.\n 2. Tons of repeat code, eg. multiple ad-hoc implementations of hash functions / PRNGs.\n 3. Almost no high-level comments about structure - I assume all that lives in some CLAUDE.md instead.Reverse-engineering through tests have never been easier, which could collapse the complexity and clean the code.","title":null,"type":"comment","url":null},{"author":"maplethorpe","children":[{"author":"bakugo","children":[{"author":"otterley","children":[{"author":"tomjakubowski","children":[],"created_at":"2026-03-31T17:45:59.000Z","created_at_i":1774979159,"id":47590986,"options":[],"parent_id":47587311,"points":null,"story_id":47584540,"text":"For what one anecdote is worth: through casual use I've found a handful of annoying UI bugs in Claude Code, and all of them were already reported on the bug tracker and either still open, or auto-closed without a real resolution.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:44:30.000Z","created_at_i":1774964670,"id":47587311,"options":[],"parent_id":47586915,"points":null,"story_id":47584540,"text":"All software that\u2019s popular has hundreds or thousands of issues filed against it. It\u2019s not an objective indication of anything other than people having issues to report and a willingness and ability to report the issue.
It doesn\u2019t mean every issue is valid, that it contains a suggestion that can be implemented, that it can be addressed immediately, etc. The issue list might not be curated, either, resulting in a garbage heap.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:15:30.000Z","created_at_i":1774962930,"id":47586915,"options":[],"parent_id":47586153,"points":null,"story_id":47584540,"text":"Nothing, apparently, which is probably why Claude Code has 7893 open issues on Github at the time of writing.","title":null,"type":"comment","url":null},{"author":"nvarsj","children":[],"created_at":"2026-03-31T13:37:08.000Z","created_at_i":1774964228,"id":47587195,"options":[],"parent_id":47586153,"points":null,"story_id":47584540,"text":"Cc is buggy as hell man. I frequently search the github for the issue I\u2019m having only to find 10 exact bugs that no one is looking at.
Obviously they don\u2019t care. Adoption is exploding. Boris brags about making 30 commits a day to the codebase.
Only will be an issue down the line when the codebase has such high entropy it takes months to add new features (maybe already there).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:08:51.000Z","created_at_i":1774958931,"id":47586153,"options":[],"parent_id":47585199,"points":null,"story_id":47584540,"text":"Well what is Anthropic doing differently to deal with this issue? Apparently they don't write any of their own code anymore, and they're doing fine.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:22:18.000Z","created_at_i":1774952538,"id":47585199,"options":[],"parent_id":47585105,"points":null,"story_id":47584540,"text":"This is the dumbest take there is about vibe coding. Claiming that managing complexity in a codebase doesn't matter anymore. I can't imagine that a competent engineer would come to the conclusion that managing complexity doesn't matter anymore. There is actually some evidence that coding agents struggle the same way humans do as the complexity of the system increases [0].
[0] https://arxiv.org/abs/2603.24755","title":null,"type":"comment","url":null},{"author":"samhh","children":[{"author":"tokioyoyo","children":[{"author":"mattmanser","children":[{"author":"tokioyoyo","children":[{"author":"mattmanser","children":[],"created_at":"2026-03-31T15:15:41.000Z","created_at_i":1774970141,"id":47588581,"options":[],"parent_id":47586812,"points":null,"story_id":47584540,"text":"Do you feel like rescinding your comment now this article is on the the HN front-page:
"Anthropic: Claude Code users hitting usage limits 'way faster than expected'"
https://news.ycombinator.com/item?id=47586176
Anthropic themselves have confirmed that something's wrong on reddit:
https://old.reddit.com/r/Anthropic/comments/1s7zfap/investig...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:07:20.000Z","created_at_i":1774962440,"id":47586812,"options":[],"parent_id":47586189,"points":null,"story_id":47584540,"text":"And those bugs were semi-fixed and people are still using it. So speed of fixes are there.
I can\u2019t comment on Claude Desktop, sorry. Personally haven\u2019t used it much.
The token usage looks like is intentional.
And I agree about the underlying model being the moat. If there\u2019s something marginally better that comes up, people will switch to it (myself included). But for now it\u2019s doing the job, despite all the hiccups, code quality and etc.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:13:04.000Z","created_at_i":1774959184,"id":47586189,"options":[],"parent_id":47585272,"points":null,"story_id":47584540,"text":"Two weeks ago typing became super laggy. It was totally unusable.
Last week I had to reinstall Claude Desktop because every time I opened it, it just hung.
This week I am sometimes opening it and getting a blank screen. It eventually works after I open it a few times.
And of course there's people complaining that somehow they're blowing their 5 hour token budget in 5 messages.
It's really buggy.
There's only so long their model will be their advantage before they all become very similar, and then the difference will be how reliable the tools are.
Right now the Claude Code code quality seems extremely low.","title":null,"type":"comment","url":null},{"author":"aerhardt","children":[],"created_at":"2026-03-31T19:11:06.000Z","created_at_i":1774984266,"id":47592043,"options":[],"parent_id":47585272,"points":null,"story_id":47584540,"text":"I've read a lot of people complain that it's buggy, here and in other forums.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:31:43.000Z","created_at_i":1774953103,"id":47585272,"options":[],"parent_id":47585220,"points":null,"story_id":47584540,"text":"Genuinely I have no idea what you mean by buggy. Sure there are some problems here and there, but my personal threshold for \u201cbuggy\u201d is much higher. I guess, for a lot of other people as well, given the uptake and usage.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:25:42.000Z","created_at_i":1774952742,"id":47585220,"options":[],"parent_id":47585105,"points":null,"story_id":47584540,"text":"It may be economically effective but such heartless, buggy software is a drain to use. I care about that delta, and yes this can be extrapolated to other industries.","title":null,"type":"comment","url":null},{"author":"ghywertelling","children":[],"created_at":"2026-03-31T12:39:28.000Z","created_at_i":1774960768,"id":47586479,"options":[],"parent_id":47585105,"points":null,"story_id":47584540,"text":"Do compilers care about their assembly generated code to look good? We will soon reach that state with all the production code. LLMs will be the compiler and actual today's human code will be replaced by LLM generated assembly code, kinda sorta human readable.","title":null,"type":"comment","url":null},{"author":"drstewart","children":[],"created_at":"2026-03-31T13:10:37.000Z","created_at_i":1774962637,"id":47586853,"options":[],"parent_id":47585105,"points":null,"story_id":47584540,"text":">Two things that matter at this day and age.
That's all that has mattered in every day and age.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:11:53.000Z","created_at_i":1774951913,"id":47585105,"options":[],"parent_id":47584709,"points":null,"story_id":47584540,"text":"It really doesn\u2019t matter anymore. I\u2019m saying this as a person who used to care about it. It does what it\u2019s generally supposed to do, it has users. Two things that matter at this day and age.","title":null,"type":"comment","url":null},{"author":"linesofcode","children":[],"created_at":"2026-03-31T11:15:10.000Z","created_at_i":1774955710,"id":47585678,"options":[],"parent_id":47584709,"points":null,"story_id":47584540,"text":"Code quality no longer carries the same weight as it did pre LLMs. It used to matter becuase humans were the ones reading/writing it so you had to optimize for readability and maintainability. But these days what matters is the AI can work with it and you can reliably test it. Obviously you don\u2019t want code quality to go totally down the drain, but there is a fine balance.
Optimize for consistency and a well thought out architecture, but let the gnarly looking function remain a gnarly function until it breaks and has to be refactored. Treat the functions as black boxes.
Personally the only time I open my IDE to look at code, it\u2019s because I\u2019m looking at something mission critical or very nuanced. For the remainder I trust my agent to deliver acceptable results.","title":null,"type":"comment","url":null},{"author":"bakugo","children":[],"created_at":"2026-03-31T13:14:00.000Z","created_at_i":1774962840,"id":47586892,"options":[],"parent_id":47584709,"points":null,"story_id":47584540,"text":"It's impressive how fast vibe coders seem to flip-flop between "AI can write better code than you, there's no reason to write code yourself anymore; if you do, you're stuck in the past" and "AI writes bad code but I don't care about quality and neither should you; if you care, you're stuck in the past".
I hope this leak can at least help silence the former. If you're going to flood the world with slop, at least own up to it.","title":null,"type":"comment","url":null},{"author":"cruffle_duffle","children":[],"created_at":"2026-03-31T19:30:57.000Z","created_at_i":1774985457,"id":47592277,"options":[],"parent_id":47584709,"points":null,"story_id":47584540,"text":"It'd dogfooding the entire concept of vibe coding and honestly, that is a good thing. Obviously they care about that stuff, but if your ethos is "always vibe code" then a lot of the fixes to it become model & prompting changes to get the thing to act like a better coder / agent / sysadmin / whatever.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:19:48.000Z","created_at_i":1774948788,"id":47584709,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"The code looks, at a glance, as bad as you expect.","title":null,"type":"comment","url":null},{"author":"bob1029","children":[{"author":"yunwal","children":[],"created_at":"2026-03-31T11:16:36.000Z","created_at_i":1774955796,"id":47585691,"options":[],"parent_id":47584711,"points":null,"story_id":47584540,"text":"> The source code of the slot machine is not relevant to the casino manager.
Famously code leaks/reverse engineering attempts of slot machines matter enormously to casino managers
[0] -https://en.wikipedia.org/wiki/Ronald_Dale_Harris#:~:text=Ron...
[1] - https://cybernews.com/news/software-glitch-loses-casino-mill...
[2] - https://sccgmanagement.com/sccg-news/2025/9/24/superbet-pays...","title":null,"type":"comment","url":null},{"author":"hmokiguess","children":[],"created_at":"2026-03-31T13:59:28.000Z","created_at_i":1774965568,"id":47587525,"options":[],"parent_id":47584711,"points":null,"story_id":47584540,"text":"That\u2019s not a good analogy, in a casino you don\u2019t own the slot machine, in this case you download the client side code to your machine","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:20:11.000Z","created_at_i":1774948811,"id":47584711,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Is this significant?
Copilot on OAI reveals everything meaningful about its functionality if you use a custom model config via the API. All you need to do is inspect the logs to see the prompts they're using. So far no one seems to care about this "loophole". Presumably, because the only thing that matters is for you to consume as many tokens per unit time as possible.
The source code of the slot machine is not relevant to the casino manager. He only cares that the customer is using it.","title":null,"type":"comment","url":null},{"author":"vbezhenar","children":[{"author":"ivanjermakov","children":[{"author":"Monotoko","children":[],"created_at":"2026-03-31T10:31:07.000Z","created_at_i":1774953067,"id":47585266,"options":[],"parent_id":47585230,"points":null,"story_id":47584540,"text":"Very easily these days, even if minified is difficult for me to reverse engineer... Claude has a very easy time of finding exactly what to patch to fix something","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:27:31.000Z","created_at_i":1774952851,"id":47585230,"options":[],"parent_id":47584778,"points":null,"story_id":47584540,"text":"> It exposes all your frontend source code for everyone
I hope it's a common knowledge that _any_ client side JavaScript is exposed to everyone. Perhaps minimized, but still easily reverse-engineerable.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:29:06.000Z","created_at_i":1774949346,"id":47584778,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"LoL! https://news.ycombinator.com/item?id=30337690
Not exactly this, but close.","title":null,"type":"comment","url":null},{"author":"avaer","children":[{"author":"NitpickLawyer","children":[{"author":"avaer","children":[{"author":"stingraycharles","children":[{"author":"falcor84","children":[],"created_at":"2026-03-31T14:33:42.000Z","created_at_i":1774967622,"id":47587966,"options":[],"parent_id":47585614,"points":null,"story_id":47584540,"text":"> detecting whether the user is using the exact same system prompt and tool definition as Claude Code
Why would it be the exact same one? Now that we have the code, it's trivial to have it randomize the prompt a bit on different requests.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:09:06.000Z","created_at_i":1774955346,"id":47585614,"options":[],"parent_id":47584851,"points":null,"story_id":47584540,"text":"I don\u2019t think that\u2019s a good comparison. There isn\u2019t anything preventing Anthropic from, say, detecting whether the user is using the exact same system prompt and tool definition as Claude Code and call it a day. Will make developing other apps nearly impossible.
It\u2019s a dynamic, subscription based service, not a static asset like a video.","title":null,"type":"comment","url":null},{"author":"esperent","children":[{"author":"raincole","children":[],"created_at":"2026-03-31T14:16:24.000Z","created_at_i":1774966584,"id":47587745,"options":[],"parent_id":47586136,"points":null,"story_id":47584540,"text":"Lol what? There is no value. OpenCode and Pi and more exist. Arguably Claude Code is the worst client on the market. People use Claude Code not because it's some amazing software. It's to access Opus at a discounted rate.","title":null,"type":"comment","url":null},{"author":"somehnguy","children":[],"created_at":"2026-03-31T15:07:55.000Z","created_at_i":1774969675,"id":47588470,"options":[],"parent_id":47586136,"points":null,"story_id":47584540,"text":"I have no interest in Claude Code as a harness, only their models. I'm used to OpenCode at this point and don't want to switch to a proprietary harness.","title":null,"type":"comment","url":null},{"author":"jen20","children":[],"created_at":"2026-03-31T17:43:56.000Z","created_at_i":1774979036,"id":47590961,"options":[],"parent_id":47586136,"points":null,"story_id":47584540,"text":"You can already do that though? [1]
[1]: https://docs.ollama.com/integrations/claude-code","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:06:21.000Z","created_at_i":1774958781,"id":47586136,"options":[],"parent_id":47584851,"points":null,"story_id":47584540,"text":"If you're using a claude subscription you'd just use claude code.
The real value here will be in using other cheap models with the cc harness.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:40:16.000Z","created_at_i":1774950016,"id":47584851,"options":[],"parent_id":47584829,"points":null,"story_id":47584540,"text":"You'd be playing cat and mouse like yt-dlp, but there's probably more value to this code than just a temporary way to milk claude subscriptions.","title":null,"type":"comment","url":null},{"author":"woleium","children":[{"author":"echelon","children":[],"created_at":"2026-03-31T10:35:06.000Z","created_at_i":1774953306,"id":47585291,"options":[],"parent_id":47584865,"points":null,"story_id":47584540,"text":""Approach Sonnet"...
So not even close to Opus, then?
These are a year behind, if not more. And they're probably clunky to use.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:42:10.000Z","created_at_i":1774950130,"id":47584865,"options":[],"parent_id":47584829,"points":null,"story_id":47584540,"text":"Just use one of the distilled claude clones instead \nhttps://x.com/0xsero/status/2038021723719688266?s=46","title":null,"type":"comment","url":null},{"author":"pkaeding","children":[{"author":"NitpickLawyer","children":[],"created_at":"2026-03-31T13:49:40.000Z","created_at_i":1774964980,"id":47587380,"options":[],"parent_id":47585207,"points":null,"story_id":47584540,"text":"Sure, but that'd be charged at API pricing. I'm talking about subscription mode above.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:23:58.000Z","created_at_i":1774952638,"id":47585207,"options":[],"parent_id":47584829,"points":null,"story_id":47584540,"text":"Could you use claude via aws bedrock?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:36:06.000Z","created_at_i":1774949766,"id":47584829,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"The problem is the oauth and their stance on bypassing that. You'd want to use your subscription, and they probably can detect that and ban users. They hold all the power there.","title":null,"type":"comment","url":null},{"author":"dahcryn","children":[{"author":"larodi","children":[],"created_at":"2026-03-31T11:10:07.000Z","created_at_i":1774955407,"id":47585622,"options":[],"parent_id":47585601,"points":null,"story_id":47584540,"text":"It reads attribution really?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:07:11.000Z","created_at_i":1774955231,"id":47585601,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"I love the irony on seeing the contribution counter at 0
Who'd have thought, the audience who doesn't want to give back to the opensource community, giving 0 contributions...","title":null,"type":"comment","url":null},{"author":"kelnos","children":[{"author":"magistr4te","children":[],"created_at":"2026-03-31T11:37:51.000Z","created_at_i":1774957071,"id":47585859,"options":[],"parent_id":47585773,"points":null,"story_id":47584540,"text":"It is a real product. They take real payments and deliver on whats promised.\nNot sure if its an attempt to subvert criticism by using satirical language, or if they truly have so little respect for the open source community.","title":null,"type":"comment","url":null},{"author":"otikik","children":[],"created_at":"2026-03-31T11:48:50.000Z","created_at_i":1774957730,"id":47585972,"options":[],"parent_id":47585773,"points":null,"story_id":47584540,"text":"Yeah... look again.
https://www.youtube.com/watch?v=6godSEVvcmU","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:25:46.000Z","created_at_i":1774956346,"id":47585773,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"Oh god, I was so close to believing Malus was a real product and not satire.","title":null,"type":"comment","url":null},{"author":"rvnx","children":[{"author":"otikik","children":[],"created_at":"2026-03-31T11:46:23.000Z","created_at_i":1774957583,"id":47585943,"options":[],"parent_id":47585820,"points":null,"story_id":47584540,"text":"They accept real money though.
https://www.youtube.com/watch?v=6godSEVvcmU","title":null,"type":"comment","url":null},{"author":"LelouBil","children":[{"author":"quadruple","children":[{"author":"LelouBil","children":[],"created_at":"2026-03-31T23:54:55.000Z","created_at_i":1775001295,"id":47595041,"options":[],"parent_id":47588238,"points":null,"story_id":47584540,"text":"That's fascinating !
I think it's worth posting as its own submission (if it wasn't already).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:53:59.000Z","created_at_i":1774968839,"id":47588238,"options":[],"parent_id":47587628,"points":null,"story_id":47584540,"text":"Semi-related, someone made basically Malus-for-San-Andreas: https://www.youtube.com/watch?v=zBQJYMKmwAs","title":null,"type":"comment","url":null},{"author":"throawayonthe","children":[{"author":"LelouBil","children":[],"created_at":"2026-03-31T23:49:11.000Z","created_at_i":1775000951,"id":47594995,"options":[],"parent_id":47588259,"points":null,"story_id":47584540,"text":"Yes, because we have clear precedent that distributing Art (the assets) is illegal by current copyright law.
But do we have precedent (in any country) that distributing different source code that compiles to the exact same binary is illegal ?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:55:55.000Z","created_at_i":1774968955,"id":47588259,"options":[],"parent_id":47587628,"points":null,"story_id":47584540,"text":"i think most game decompilation projects are either openly illegal or operate on "provide your own binary" and build automatic tooling around it","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:08:08.000Z","created_at_i":1774966088,"id":47587628,"options":[],"parent_id":47585820,"points":null,"story_id":47584540,"text":"First time I hear about this, it's interesting to have written all of this out.
Now this makes me think of game decompilation projects, which would seem to fall in the same legal area as code that would be generated by something like Malus.
Different code, same end result (binary or api).
We definitely need to know what the legal limits are and should be","title":null,"type":"comment","url":null},{"author":"chillfox","children":[{"author":"monooso","children":[{"author":"tomjakubowski","children":[],"created_at":"2026-03-31T17:32:30.000Z","created_at_i":1774978350,"id":47590790,"options":[],"parent_id":47588454,"points":null,"story_id":47584540,"text":"And we know they're right, because that lawyer signed a contract on TV saying he'd be liable if they were wrong.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:07:15.000Z","created_at_i":1774969635,"id":47588454,"options":[],"parent_id":47587891,"points":null,"story_id":47584540,"text":"Dumb Starbucks begs to differ.
https://en.wikipedia.org/wiki/Dumb_Starbucks","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:27:31.000Z","created_at_i":1774967251,"id":47587891,"options":[],"parent_id":47585820,"points":null,"story_id":47584540,"text":"It's not a parody when they accept money and deliver the service.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:32:36.000Z","created_at_i":1774956756,"id":47585820,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"Malus is not a real project btw, it's a parody:
\u201cLet's end open source together with this one simple trick\u201d
https://pretalx.fosdem.org/fosdem-2026/talk/SUVS7G/feedback/
Malus is translating code into text, and from text back into code.
It gives the illusion of clean room implementation that some companies abuse.
The irony is that ChatGPT/Claude answers are all actually directly derived from open-source code, so...","title":null,"type":"comment","url":null},{"author":"gosub100","children":[],"created_at":"2026-03-31T12:03:28.000Z","created_at_i":1774958608,"id":47586108,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"What are they worried about? Someone taking the company's job? Hehe","title":null,"type":"comment","url":null},{"author":"aizk","children":[],"created_at":"2026-03-31T12:21:16.000Z","created_at_i":1774959676,"id":47586268,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"This has happened before. It was called anon kode.","title":null,"type":"comment","url":null},{"author":"TIPSIO","children":[],"created_at":"2026-03-31T12:37:22.000Z","created_at_i":1774960642,"id":47586456,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"Eh, the value is the unlimited Max plan which they have rightfully banned from third-party use.
People simply want Opus without fear of billing nightmare.
That\u2019s like 99% of it.","title":null,"type":"comment","url":null},{"author":"sumeno","children":[{"author":"williamcotton","children":[{"author":"fsmv","children":[],"created_at":"2026-03-31T14:56:25.000Z","created_at_i":1774968985,"id":47588267,"options":[],"parent_id":47587814,"points":null,"story_id":47584540,"text":"Trade secrets once made public don't have any legal protection and I haven't signed any contract with anthropic","title":null,"type":"comment","url":null},{"author":"jpetso","children":[{"author":"bigbuppo","children":[],"created_at":"2026-03-31T17:47:09.000Z","created_at_i":1774979229,"id":47591003,"options":[],"parent_id":47588395,"points":null,"story_id":47584540,"text":"Yeah, but you don't have trillions of dollars of investments riding on your success, so the rules still apply to you.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:03:36.000Z","created_at_i":1774969416,"id":47588395,"options":[],"parent_id":47587814,"points":null,"story_id":47584540,"text":"Apparently it's possible to download a whole load of books illegally, but still train AI models on them without those getting pulled after you get found out.
The same reasoning may apply here :P","title":null,"type":"comment","url":null},{"author":"dns_snek","children":[],"created_at":"2026-03-31T15:18:51.000Z","created_at_i":1774970331,"id":47588635,"options":[],"parent_id":47587814,"points":null,"story_id":47584540,"text":"They published the code on their own, none of that applies.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:22:18.000Z","created_at_i":1774966938,"id":47587814,"options":[],"parent_id":47586492,"points":null,"story_id":47584540,"text":"What about trade secrets, breach of contract, etc, etc?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:40:30.000Z","created_at_i":1774960830,"id":47586492,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"No real reason to do that, they say Claude Code is written by Claude, which means it has no copyright. Just use the code directly","title":null,"type":"comment","url":null},{"author":"conradfr","children":[],"created_at":"2026-03-31T18:23:36.000Z","created_at_i":1774981416,"id":47591463,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"Maybe https://github.com/instructkr/claw-code","title":null,"type":"comment","url":null},{"author":"DamnInteresting","children":[],"created_at":"2026-04-01T02:21:37.000Z","created_at_i":1775010097,"id":47596018,"options":[],"parent_id":47584821,"points":null,"story_id":47584540,"text":"Malus is a homophone of "malice." But then again "Claude" is a homophone of "clod."","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:34:50.000Z","created_at_i":1774949690,"id":47584821,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Would be interesting to run this through Malus [1] or literally just Claude Code and get open source Claude Code out of it.
I jest, but in a world where these models have been trained on gigatons of open source I don't even see the moral problem. IANAL, don't actually do this.
https://malus.sh/","title":null,"type":"comment","url":null},{"author":"karimf","children":[{"author":"simonklee","children":[],"created_at":"2026-03-31T11:38:37.000Z","created_at_i":1774957117,"id":47585867,"options":[],"parent_id":47584882,"points":null,"story_id":47584540,"text":"Not really, except that they have a bunch of weird things in the source code and people like to make fun of it. OpenCode/Codex generally doesn't have this since these are open-source projects from the get go.
(I work on OpenCode)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:46:07.000Z","created_at_i":1774950367,"id":47584882,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Is there anything special here vs. OpenCode or Codex?
There were/are a lot of discussions on how the harness can affect the output.","title":null,"type":"comment","url":null},{"author":"bryanhogan","children":[{"author":"dang","children":[],"created_at":"2026-03-31T17:47:28.000Z","created_at_i":1774979248,"id":47591011,"options":[],"parent_id":47584945,"points":null,"story_id":47584540,"text":"Added to toptext. Thanks!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:53:45.000Z","created_at_i":1774950825,"id":47584945,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"https://xcancel.com/Fried_rice/status/2038894956459290963","title":null,"type":"comment","url":null},{"author":"dheerajmp","children":[{"author":"zhisme","children":[{"author":"treexs","children":[{"author":"panny","children":[{"author":"krlx","children":[{"author":"shimman","children":[],"created_at":"2026-03-31T14:03:40.000Z","created_at_i":1774965820,"id":47587566,"options":[],"parent_id":47585580,"points":null,"story_id":47584540,"text":"It's undetermined if code will be majority written by machines, especially as people start to realize how harmful these tools are without extreme diligence. Outages at Cloudflare, AWS, GitHub, etc are just the beginning. Companies aren't going to want to use tools that can potentially cause $100s of millions in potential damages (see Amazon store being down causing massive revenue loss).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:05:06.000Z","created_at_i":1774955106,"id":47585580,"options":[],"parent_id":47585420,"points":null,"story_id":47584540,"text":"Given that from 2026 onwards most of the code is going to be computer generated, doesn't it open some interesting implications there ?","title":null,"type":"comment","url":null},{"author":"0x3f","children":[{"author":"panny","children":[],"created_at":"2026-03-31T12:06:44.000Z","created_at_i":1774958804,"id":47586139,"options":[],"parent_id":47585757,"points":null,"story_id":47584540,"text":"That's fine. File a false claim DMCA and that's felony perjury :) They know for a fact that there is no copyright on AI generated code, the courts have affirmed this repeatedly.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:24:14.000Z","created_at_i":1774956254,"id":47585757,"options":[],"parent_id":47585420,"points":null,"story_id":47584540,"text":"I'm sure it's not _entirely_ built that way, and in practically speaking GitHub will almost certainly take it down rather than doing some kind of deep research about which code is which.","title":null,"type":"comment","url":null},{"author":"nananana9","children":[],"created_at":"2026-03-31T12:25:30.000Z","created_at_i":1774959930,"id":47586311,"options":[],"parent_id":47585420,"points":null,"story_id":47584540,"text":"Try not to be overly confident about things where even the experts in the field (copyright lawyers) are uncertain of.
There's no major lawsuits about this yet, the general consensus is that even under current regulations it's in the grey. And even if you turn out to be right, and let's say 99% of this code is AI-generated, you're still breaking the law by using the other 1%, and good luck proving in court what parts of their code were human written and what weren't (especially when being sued by the company that literally has the LLM logs).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:48:45.000Z","created_at_i":1774954125,"id":47585420,"options":[],"parent_id":47585333,"points":null,"story_id":47584540,"text":"They can't. AI generated code cannot be copyrighted. They've stated that claude code is built with claude code. You can take this and start your own claude code project now if you like. There's zero copyright protection on this.","title":null,"type":"comment","url":null},{"author":"paxys","children":[],"created_at":"2026-03-31T11:52:38.000Z","created_at_i":1774957958,"id":47586014,"options":[],"parent_id":47585333,"points":null,"story_id":47584540,"text":"Which is why you should clone it right now","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:40:03.000Z","created_at_i":1774953603,"id":47585333,"options":[],"parent_id":47585322,"points":null,"story_id":47584540,"text":"won't they just try to dmca or take these down especially if they're more popular","title":null,"type":"comment","url":null},{"author":"moontear","children":[{"author":"nubinetwork","children":[],"created_at":"2026-03-31T12:25:03.000Z","created_at_i":1774959903,"id":47586307,"options":[],"parent_id":47586184,"points":null,"story_id":47584540,"text":"Looks like mostly spam making fun of the code leak.","title":null,"type":"comment","url":null},{"author":"sudo_man","children":[],"created_at":"2026-03-31T13:10:48.000Z","created_at_i":1774962648,"id":47586858,"options":[],"parent_id":47586184,"points":null,"story_id":47584540,"text":"too much wechat QR Codes","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:12:28.000Z","created_at_i":1774959148,"id":47586184,"options":[],"parent_id":47585322,"points":null,"story_id":47584540,"text":"Popular, yes... but have you seen the issues? SOMETHING is going on in that repo: https://github.com/instructkr/claude-code/issues","title":null,"type":"comment","url":null},{"author":"101008","children":[],"created_at":"2026-03-31T12:57:11.000Z","created_at_i":1774961831,"id":47586700,"options":[],"parent_id":47585322,"points":null,"story_id":47584540,"text":"which has already been deleted","title":null,"type":"comment","url":null},{"author":"ezekg","children":[{"author":"BoorishBears","children":[{"author":"ezekg","children":[{"author":"BoorishBears","children":[],"created_at":"2026-04-01T17:14:40.000Z","created_at_i":1775063680,"id":47603704,"options":[],"parent_id":47596544,"points":null,"story_id":47584540,"text":"So am I.
That project you quoted is the one with that as its new description. Soon it'll just be [new thing] that happens to use the stars as social proof... in fact when I look again:
> The fastest repo in history to surpass 100K stars . Better Harness Tools that make real things done. Built in Rust using oh-my-codex.
They started a new project that justifies the same repo and scrapes a little credibility off of Claude Code. The intent is not an actual rewrite but to bolster what will be their own personal project trying to compete with OpenCode and co.
The grifter is already pasting references to WSJ articles about themselves in the Readme","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T03:41:35.000Z","created_at_i":1775014895,"id":47596544,"options":[],"parent_id":47595180,"points":null,"story_id":47584540,"text":"> The result is a clean-room Python rewrite that captures the architectural patterns of Claude Code's agent harness without copying any proprietary source.
This is what I'm referring to.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T00:12:49.000Z","created_at_i":1775002369,"id":47595180,"options":[],"parent_id":47589151,"points":null,"story_id":47584540,"text":"That's not the actual plan.
"I have a popular repo, but the content will likely be removed and I won't have personally gained from the saga: how can I fix the part where I didn't profit?"
"Eureka! I'll remove the content preemptively, then come up with a backstory that justifies reusing the now empty repo for building the umpteenth coding harness! And I can even claim fuzzy ties to Claude Code!"
Hence the new description:
> The fastest repo in history to surpass 50K stars , reaching the milestone in just 2 hours after publication. Better Harness Tools, not merely storing the archive of leaked Claude Code but also make real things done. Now rewriting in Rust.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:47:39.000Z","created_at_i":1774972059,"id":47589151,"options":[],"parent_id":47585322,"points":null,"story_id":47584540,"text":"I don't understand how you can have a 'clean-room port.' Seems contradictory to me.","title":null,"type":"comment","url":null},{"author":"DrammBA","children":[],"created_at":"2026-03-31T16:23:42.000Z","created_at_i":1774974222,"id":47589761,"options":[],"parent_id":47585322,"points":null,"story_id":47584540,"text":"What do stars mean in the context of random github accounts mirroring leaked source code?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:38:35.000Z","created_at_i":1774953515,"id":47585322,"options":[],"parent_id":47584963,"points":null,"story_id":47584540,"text":"https://github.com/instructkr/claude-code
this one has more stars and more popular","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:56:42.000Z","created_at_i":1774951002,"id":47584963,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Source here https://github.com/chatgptprojects/claude-code/","title":null,"type":"comment","url":null},{"author":"Squarex","children":[{"author":"rafram","children":[{"author":"loveparade","children":[],"created_at":"2026-03-31T10:22:14.000Z","created_at_i":1774952534,"id":47585197,"options":[],"parent_id":47585078,"points":null,"story_id":47584540,"text":"Your favorite LLM is great at building a super buggy renderer, so that's no longer a moat","title":null,"type":"comment","url":null},{"author":"rick_dalton","children":[],"created_at":"2026-03-31T19:00:40.000Z","created_at_i":1774983640,"id":47591913,"options":[],"parent_id":47585078,"points":null,"story_id":47584540,"text":"Gemini-cli is much worse in my experience but I agree","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:09:03.000Z","created_at_i":1774951743,"id":47585078,"options":[],"parent_id":47585001,"points":null,"story_id":47584540,"text":"Well, Claude does boast an absolutely cursed (and very buggy) React-based TUI renderer that I think the others lack! What if someone steals it and builds their own buggy TUI app?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:01:50.000Z","created_at_i":1774951310,"id":47585001,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Codex and gemini cli are open source already. And plenty of other agents. I don't think there is any moat in claude code source.","title":null,"type":"comment","url":null},{"author":"theanonymousone","children":[{"author":"treexs","children":[{"author":"theanonymousone","children":[],"created_at":"2026-03-31T11:02:41.000Z","created_at_i":1774954961,"id":47585563,"options":[],"parent_id":47585347,"points":null,"story_id":47584540,"text":"I believe GitHub can and does suspend accounts that use such proxies.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:41:37.000Z","created_at_i":1774953697,"id":47585347,"options":[],"parent_id":47585010,"points":null,"story_id":47584540,"text":"does this not work? https://www.mintlify.com/samarth777/claude-code-copilot/intr...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:02:50.000Z","created_at_i":1774951370,"id":47585010,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I am waiting now for someone to make it work with a Copilot Pro subscription.","title":null,"type":"comment","url":null},{"author":"anhldbk","children":[{"author":"DeathArrow","children":[],"created_at":"2026-03-31T10:09:45.000Z","created_at_i":1774951785,"id":47585086,"options":[],"parent_id":47585026,"points":null,"story_id":47584540,"text":"And while they are at it, open source Opus and Sonet. :)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:04:00.000Z","created_at_i":1774951440,"id":47585026,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I guess it's time for Anthropic to open source Claude Code.","title":null,"type":"comment","url":null},{"author":"lukan","children":[{"author":"mikrotikker","children":[{"author":"singularity2001","children":[{"author":"prmoustache","children":[],"created_at":"2026-03-31T11:01:28.000Z","created_at_i":1774954888,"id":47585547,"options":[],"parent_id":47585382,"points":null,"story_id":47584540,"text":"This is neither.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:44:25.000Z","created_at_i":1774953865,"id":47585382,"options":[],"parent_id":47585338,"points":null,"story_id":47584540,"text":"You are not allowed to use the assistance of Claude to manufacture hacks and bombs on your computer","title":null,"type":"comment","url":null},{"author":"lukan","children":[{"author":"mattmanser","children":[{"author":"lukan","children":[],"created_at":"2026-03-31T12:08:39.000Z","created_at_i":1774958919,"id":47586151,"options":[],"parent_id":47586109,"points":null,"story_id":47584540,"text":"I know. All I could do in that case is a blogpost "Claude banned me, for following claude's instructions!" and hope it gets viral.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:03:30.000Z","created_at_i":1774958610,"id":47586109,"options":[],"parent_id":47585417,"points":null,"story_id":47584540,"text":"It doesn't matter what defence you can think of, if they want to ban you, they'll ban you.
They won't even read your defence.","title":null,"type":"comment","url":null},{"author":"dheera","children":[],"created_at":"2026-03-31T20:01:02.000Z","created_at_i":1774987262,"id":47592669,"options":[],"parent_id":47585417,"points":null,"story_id":47584540,"text":"> I do not like to get banned
This is why I do such experiments on ChatGPT and not Claude.
I don't want to get banned by Claude but I couldn't care less if ChatGPT bans me.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:48:34.000Z","created_at_i":1774954114,"id":47585417,"options":[],"parent_id":47585338,"points":null,"story_id":47584540,"text":"Well, the thing is, I do not just use my computer, but connect to their computers and I do not like to get banned. I suppose simple UI things like expanding source files won't change a thing, but the more interesting things, editing the context etc. do have that risk, but no idea if they look for it or enforce it. Their side is, if I want to have full control, I need to use the API directly(way more expensive) and what I want to do is basically circumventing it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:40:36.000Z","created_at_i":1774953636,"id":47585338,"options":[],"parent_id":47585039,"points":null,"story_id":47584540,"text":"I think that you do not need to feel uneasy at all. It is your computer and your memory space that the data is stored and operating in you can do whatever you like to the bits in that space. I would encourage you to continue that experiment.","title":null,"type":"comment","url":null},{"author":"bredren","children":[{"author":"lukan","children":[],"created_at":"2026-03-31T20:13:06.000Z","created_at_i":1774987986,"id":47592818,"options":[],"parent_id":47592384,"points":null,"story_id":47584540,"text":"The first trick will be avoiding getting flagged for running an unofficial build.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:39:43.000Z","created_at_i":1774985983,"id":47592384,"options":[],"parent_id":47585039,"points":null,"story_id":47584540,"text":"The trick isn't to patch it once, but to create a system that can reproduce your patches against each release as they come in. Then, when code changes make fixes non-trivial calling in a headless session to heal your fixes.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:04:49.000Z","created_at_i":1774951489,"id":47585039,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Neat. Coincidently recently I asked Claude about Claude CLI, if it is possible to patch some annoying things (like not being able to expand Ctrl + O more than once, so never be able to see some lines and in general have more control over the context) and it happily proclaimed it is open source and it can do it ... and started doing something. Then I checked a bit and saw, nope, not open source. And by the wording of the TOS, it might brake some sources. But claude said, "no worries", it only break the TOS technically. So by saving that conversation I would have some defense if I would start messing with it, but felt a bit uneasy and stopped the experiment. Also claude came into a loop, but if I would point it at this, it might work I suppose.","title":null,"type":"comment","url":null},{"author":"DeathArrow","children":[{"author":"epolanski","children":[],"created_at":"2026-03-31T10:09:28.000Z","created_at_i":1774951768,"id":47585081,"options":[],"parent_id":47585044,"points":null,"story_id":47584540,"text":"Responsibility goes upwards.
Why weren't proper checks in place in the first place?
Bonus: why didn't they setup their own AI-assisted tools to harness the release checks?","title":null,"type":"comment","url":null},{"author":"matltc","children":[],"created_at":"2026-03-31T10:14:54.000Z","created_at_i":1774952094,"id":47585134,"options":[],"parent_id":47585044,"points":null,"story_id":47584540,"text":"Ha. I'm surprised it's not a CI job","title":null,"type":"comment","url":null},{"author":"orphea","children":[],"created_at":"2026-03-31T12:24:20.000Z","created_at_i":1774959860,"id":47586303,"options":[],"parent_id":47585044,"points":null,"story_id":47584540,"text":"
the poor guy\nOriginal llama models leaked from meta. Instead of fighting it they decided to publish them officially. Real boost to the OS/OW models movement, they have been leading it for a while after that.
It would be interesting to see that same thing with CC, but I doubt it'll ever happen.","title":null,"type":"comment","url":null},{"author":"ChicagoDave","children":[],"created_at":"2026-03-31T10:11:06.000Z","created_at_i":1774951866,"id":47585096,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I hope everyone provides excellent feedback so they improve Claude Code.","title":null,"type":"comment","url":null},{"author":"DeathArrow","children":[{"author":"jsk2600","children":[{"author":"ghywertelling","children":[{"author":"arthur-st","children":[],"created_at":"2026-03-31T13:04:27.000Z","created_at_i":1774962267,"id":47586776,"options":[],"parent_id":47586501,"points":null,"story_id":47584540,"text":"Yes, that's essentially the only practical reason.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:41:55.000Z","created_at_i":1774960915,"id":47586501,"options":[],"parent_id":47585168,"points":null,"story_id":47584540,"text":"is that the reason why Anthropic acquired Bun, a javascript tooling company?","title":null,"type":"comment","url":null},{"author":"pipes","children":[],"created_at":"2026-04-09T18:41:26.000Z","created_at_i":1775760086,"id":47707835,"options":[],"parent_id":47585168,"points":null,"story_id":47584540,"text":"He commented here about it (sort of)
https://news.ycombinator.com/item?id=21934336
Thanks hackernewsbooks for sending me that!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:18:27.000Z","created_at_i":1774952307,"id":47585168,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"Original author of Claude Code is expert on TypeScript [1]
[1] https://www.amazon.com/Programming-TypeScript-Making-JavaScr...","title":null,"type":"comment","url":null},{"author":"bigbezet","children":[],"created_at":"2026-03-31T10:19:15.000Z","created_at_i":1774952355,"id":47585174,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"It's not a desktop tool, it's a CLI tool.
But a lot of desktop tools are written in JS because it's easy to create multi-platform applications.","title":null,"type":"comment","url":null},{"author":"ivanjermakov","children":[],"created_at":"2026-03-31T10:29:20.000Z","created_at_i":1774952960,"id":47585248,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"Because it's the most popular programming language in the world?","title":null,"type":"comment","url":null},{"author":"TiredOfLife","children":[],"created_at":"2026-03-31T11:25:24.000Z","created_at_i":1774956324,"id":47585772,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"I am happy you woke up from your 10 year coma.","title":null,"type":"comment","url":null},{"author":"wanttosaythings","children":[{"author":"c0wb0yc0d3r","children":[],"created_at":"2026-03-31T12:20:17.000Z","created_at_i":1774959617,"id":47586259,"options":[],"parent_id":47585877,"points":null,"story_id":47584540,"text":"This is a common take but language servers bridge the gap well.
Language servers, however, are a pain on Claude code. https://github.com/anthropics/claude-code/issues/15619","title":null,"type":"comment","url":null},{"author":"rvz","children":[],"created_at":"2026-03-31T16:15:32.000Z","created_at_i":1774973732,"id":47589628,"options":[],"parent_id":47585877,"points":null,"story_id":47584540,"text":"Would have believed you if you have said that a day later.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:40:18.000Z","created_at_i":1774957218,"id":47585877,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"LLMs are good in JS and Python which means everything from now on will be written in or ported to either of those two languages.\nSo yeah, JS is the future of all software.","title":null,"type":"comment","url":null},{"author":"progx","children":[],"created_at":"2026-03-31T12:03:49.000Z","created_at_i":1774958629,"id":47586114,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"Anthropic acquired bun last year https://bun.com/blog/bun-joins-anthropic","title":null,"type":"comment","url":null},{"author":"monkpit","children":[],"created_at":"2026-03-31T12:34:40.000Z","created_at_i":1774960480,"id":47586429,"options":[],"parent_id":47585110,"points":null,"story_id":47584540,"text":"Alternatively: why not?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:12:17.000Z","created_at_i":1774951937,"id":47585110,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Why is Claude Code, a desktop tool, written in JS? Is the future of all software JS or Typescript?","title":null,"type":"comment","url":null},{"author":"isodev","children":[{"author":"bmitc","children":[],"created_at":"2026-03-31T10:40:15.000Z","created_at_i":1774953615,"id":47585335,"options":[],"parent_id":47585143,"points":null,"story_id":47584540,"text":"The source is linked to in this thread. Is that not the source code?","title":null,"type":"comment","url":null},{"author":"echelon","children":[],"created_at":"2026-03-31T10:40:37.000Z","created_at_i":1774953637,"id":47585339,"options":[],"parent_id":47585143,"points":null,"story_id":47584540,"text":"The only exciting leak would be the Opus weights themselves.","title":null,"type":"comment","url":null},{"author":"kelnos","children":[],"created_at":"2026-03-31T11:35:30.000Z","created_at_i":1774956930,"id":47585844,"options":[],"parent_id":47585143,"points":null,"story_id":47584540,"text":"If it was included unintentionally, then it's a leak.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:15:51.000Z","created_at_i":1774952151,"id":47585143,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Can we stop referring to source maps as leaks? It was packaged in a way that wasn\u2019t even obfuscated. Same as websites - it\u2019s not a \u201cleak\u201d that you can read or inspect the source code.","title":null,"type":"comment","url":null},{"author":"cedws","children":[{"author":"nialse","children":[{"author":"petcat","children":[{"author":"arcfour","children":[{"author":"jtbayly","children":[],"created_at":"2026-03-31T13:55:56.000Z","created_at_i":1774965356,"id":47587457,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"Wut?They did exactly the same thing!
Try this: If you want to train a model, you\u2019re free to write your own books and websites to feed into it. You\u2019re not free to let others do that work for you because they don\u2019t want you to, because it cost them a lot of time and money and secret sauce presumably filtering it for quality and other stuff.","title":null,"type":"comment","url":null},{"author":"petcat","children":[{"author":"gruez","children":[{"author":"petcat","children":[{"author":"dpark","children":[{"author":"petcat","children":[{"author":"dpark","children":[],"created_at":"2026-03-31T16:30:40.000Z","created_at_i":1774974640,"id":47589872,"options":[],"parent_id":47589635,"points":null,"story_id":47584540,"text":"I didn\u2019t say anything about the computer fraud and abuse act. TOS are legally binding contracts in their own right if implemented correctly.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:16:01.000Z","created_at_i":1774973761,"id":47589635,"options":[],"parent_id":47589296,"points":null,"story_id":47584540,"text":"The Supreme Court already ruled on this. Scraping public data, or data that you are authorized to access, is not a violation of the Computer Fraud and Abuse Act.
Now, if you try to get around attempts to block your access, then yes you could be in legal trouble. But that's not what is happening here. These are people/companies that have Claude accounts in good standing and are authorized by Anthropic to access the data.
Nobody is saying that Anthropic can't just block them though, and they are certainly trying.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:56:10.000Z","created_at_i":1774972570,"id":47589296,"options":[],"parent_id":47589099,"points":null,"story_id":47584540,"text":"> They just can't legally stop you by, for instance, compelling a judge to order you to stop.
They probably can, actually. TOS are legally binding.
More likely they would block you rather than pursuing legal avenues but they certainly could.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:44:53.000Z","created_at_i":1774971893,"id":47589099,"options":[],"parent_id":47588871,"points":null,"story_id":47584540,"text":"Nobody is saying they can't try to stop you themselves. That's where the Terms of Service violation part comes in. They can cancel your account, block your IP, etc. They just can't legally stop you by, for instance, compelling a judge to order you to stop.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:32:17.000Z","created_at_i":1774971137,"id":47588871,"options":[],"parent_id":47587513,"points":null,"story_id":47584540,"text":">The courts have ruled that AI outputs are not copyrightable.
"not copyrightable" doesn't imply they can't frustrate attempts to scrape data.","title":null,"type":"comment","url":null},{"author":"dpark","children":[],"created_at":"2026-03-31T15:41:45.000Z","created_at_i":1774971705,"id":47589044,"options":[],"parent_id":47587513,"points":null,"story_id":47584540,"text":"> 2) Scraping Anthropic's model outputs is no different than what Anthropic already did. Only a TOS violation.
Regardless of whether LLM training amounts to theft, thieves are still allowed to put locks on their own doors.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:58:53.000Z","created_at_i":1774965533,"id":47587513,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"> Stole? Courts have ruled it's transformative, and it very obviously is.
The courts have ruled that AI outputs are not copyrightable. The courts have also ruled that scraping by itself is not illegal, only maybe against a Terms of Service. Therefore, Anthropic, OpenAI, Google, etc. have no legal claim to any proprietary protections of their model outputs.
So we have two things that are true:
1) Anthropic (certainly) violated numerous TOS by scraping all of the internet, not just public content.
2) Scraping Anthropic's model outputs is no different than what Anthropic already did. Only a TOS violation.","title":null,"type":"comment","url":null},{"author":"airstrike","children":[],"created_at":"2026-03-31T14:13:38.000Z","created_at_i":1774966418,"id":47587711,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"Guess who else spent a lot of time and money and secret sauce?
Do you hear the words coming out of your mouth?","title":null,"type":"comment","url":null},{"author":"unethical_ban","children":[{"author":"p1esk","children":[],"created_at":"2026-03-31T14:38:41.000Z","created_at_i":1774967921,"id":47588024,"options":[],"parent_id":47587749,"points":null,"story_id":47584540,"text":"I don\u2019t see why it\u2019s not ok to do that to an AI model. Or are you asking why they don\u2019t want you to do it?","title":null,"type":"comment","url":null},{"author":"sfn42","children":[],"created_at":"2026-03-31T14:49:19.000Z","created_at_i":1774968559,"id":47588174,"options":[],"parent_id":47587749,"points":null,"story_id":47584540,"text":"I don't think anyone's saying it's not okay - I think the point is that Anthropic has every right to create safeguards against it if they want to - just like the people publishing other information are free to do the same.
And everyone is free to consume all the free information.","title":null,"type":"comment","url":null},{"author":"gruez","children":[],"created_at":"2026-03-31T15:48:07.000Z","created_at_i":1774972087,"id":47589156,"options":[],"parent_id":47587749,"points":null,"story_id":47584540,"text":">Why is it okay for these companies to pirate books
Courts have ruled it's not, and I don't think anyone is arguing it's okay.
>but it is not okay for others to try to do the same to an AI model?
The steelman version is that it's okay to do it once you acquired the data somehow, but that doesn't mean anthropic can't set up roadblocks to frustrate you.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:16:38.000Z","created_at_i":1774966598,"id":47587749,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"Let's talk ethics, not law. Why is it okay for these companies to pirate books and scrape the entire web and offer synthesized summaries of all of it, lowering traffic and revenue for countless websites and professions of experts, but it is not okay for others to try to do the same to an AI model?
Is the work of others less valid than the work of a model?","title":null,"type":"comment","url":null},{"author":"two_tasty","children":[{"author":"kspacewalk2","children":[{"author":"salawat","children":[{"author":"dpark","children":[],"created_at":"2026-03-31T15:48:20.000Z","created_at_i":1774972100,"id":47589162,"options":[],"parent_id":47589049,"points":null,"story_id":47584540,"text":"You explained what \u201crich and poor are equally forbidden from sleeping under bridges\u201d means, but not what this has to do with the statement that one is free to do their own scraping and training, which I\u2019m pretty sure is what kspacewalk was asking.","title":null,"type":"comment","url":null},{"author":"kspacewalk2","children":[],"created_at":"2026-03-31T16:30:06.000Z","created_at_i":1774974606,"id":47589864,"options":[],"parent_id":47589049,"points":null,"story_id":47584540,"text":"Okay, you explained what Anatole France meant, which is probably helpful for those few who didn't get it from the quote itself. Perhaps now you can explain what on earth this has to do with Anthropic not wanting to let other for-profit businesses mooch off its investment of time, brainpower and money?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:42:08.000Z","created_at_i":1774971728,"id":47589049,"options":[],"parent_id":47588331,"points":null,"story_id":47584540,"text":"Rich people aren't going to find themselves needing to sleep under a bridge, so the law really only exists as a constraint on the poor. Duh. The flex that "well a rich guy couldn't do it either" is A) at best a myopic misunderstanding perpetuated by out of touch people and B) hopelessly naive, because anny punishment for the rich guy actually sleeping under a bridge is so laughably small it may as well not even exist. Hence, the whole bit of "a legal system to keep these accountable, but not for me".","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:00:29.000Z","created_at_i":1774969229,"id":47588331,"options":[],"parent_id":47588254,"points":null,"story_id":47584540,"text":"Meaning what? The poor gets to sleep in the guest room of the rich guy's house because muh inequality?
Anthropic paid a lot of money for a moat and want to guard it. It is not wrong, in any sense of the word, for them to do so.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:54:54.000Z","created_at_i":1774968894,"id":47588254,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":""...free to scrape the web yourself and train your own model."
Yes, rich and poor are equally forbidden from sleeping under bridges.","title":null,"type":"comment","url":null},{"author":"nunez","children":[{"author":"nunez","children":[],"created_at":"2026-03-31T17:31:30.000Z","created_at_i":1774978290,"id":47590771,"options":[],"parent_id":47588290,"points":null,"story_id":47584540,"text":"(429'ed, I meant)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:57:52.000Z","created_at_i":1774969072,"id":47588290,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"Lol; like heck we are. Try scraping the NYTimes at LLM scale. You can time how quickly you\u2019ll get 420\u2019ed or, at worst, hit with a C&D.","title":null,"type":"comment","url":null},{"author":"alpha_squared","children":[],"created_at":"2026-03-31T15:02:24.000Z","created_at_i":1774969344,"id":47588374,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"> You're perfectly free to scrape the web yourself and train your own model.
Actually, not anymore as a result of OpenAI and Anthropic's scraping. For example, Reddit came down hard on access to their APIs as a response to ChatGPT's release and the news that LLMs were built atop of scraping the open web. Most of the web today is not as open as before as a result of scraping for LLM data. So, no, no one is perfectly free to scrape the web anymore because open access is dying.","title":null,"type":"comment","url":null},{"author":"andersonpico","children":[],"created_at":"2026-03-31T15:45:39.000Z","created_at_i":1774971939,"id":47589117,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"Your selective respect for work is a glaring double standard. The effort to produce the original content they scraped is order of magnitudes bigger than what it took to train the model, so if this wasn't enough to protect the authors from Anthropic it shouldn't be enough to protected Anthropic from people distillating their models.
Your legal argument is all over the place as well. What is more relevant here: what the courts ruled or what you consider obvious? How is distillation less transformative than scraping? How does courts ruling that scraping to train models is legal relate to distillation?
Nobody is scoring you on neutrality points for not using AI much and calling this doomerism is just a thought-terminating cliche that refuses to engage with the comment you're replying.
In fact, your comment is not engaging with anything at all, you're vaguely gesturing towards potentitial arguments without making them. If you find discussing this exhausting then don't but also don't flood the comments with low effort whining.","title":null,"type":"comment","url":null},{"author":"loremium","children":[],"created_at":"2026-03-31T16:14:31.000Z","created_at_i":1774973671,"id":47589609,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"reminds me of `don't look up` a bit. there clearly is an imbalance in regards to licenses with model providers, not even talking about knowledge extraction (yes younger people don't learn properly now, older generations forget) shortly before the rug-pull happens in form of accessibility to not rich people","title":null,"type":"comment","url":null},{"author":"hax0ron3","children":[],"created_at":"2026-03-31T21:01:56.000Z","created_at_i":1774990916,"id":47593429,"options":[],"parent_id":47587358,"points":null,"story_id":47584540,"text":"It is transformative, but if I make a bunch of requests to their API and use the responses to distill my own model, that is also transformative.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:47:57.000Z","created_at_i":1774964877,"id":47587358,"options":[],"parent_id":47586737,"points":null,"story_id":47584540,"text":"You're perfectly free to scrape the web yourself and train your own model. You're not free to let Anthropic do that work for you, because they don't want you to, because it cost them a lot of time and money and secret sauce presumably filtering it for quality and other stuff.
Stole? Courts have ruled it's transformative, and it very obviously is.
AI doomerism is exhausting, and I don't even use AI that much, it's just annoying to see people who want to find any reason they can to moan.","title":null,"type":"comment","url":null},{"author":"olalonde","children":[],"created_at":"2026-03-31T14:24:29.000Z","created_at_i":1774967069,"id":47587852,"options":[],"parent_id":47586737,"points":null,"story_id":47584540,"text":"Also, begging to get "regulated":
https://x.com/TheChiefNerd/status/2038565951268946021","title":null,"type":"comment","url":null},{"author":"Andrex","children":[],"created_at":"2026-03-31T15:23:19.000Z","created_at_i":1774970599,"id":47588705,"options":[],"parent_id":47586737,"points":null,"story_id":47584540,"text":"I just rewatched that scene last night on YouTube. Maybe this is the universe telling me to watch the whole movie again...
It's cool to see Noah Wyle getting his due these days (The Pitt).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:00:24.000Z","created_at_i":1774962024,"id":47586737,"options":[],"parent_id":47586633,"points":null,"story_id":47584540,"text":"They stole everything and now they want to close the gates behind them.
"I got the loot, Steve!"
I feel like the distillation stuff will end up in court if they try to sue an American company about it. We'll see what a judge says.","title":null,"type":"comment","url":null},{"author":"spiderfarmer","children":[{"author":"croes","children":[{"author":"dmix","children":[{"author":"jitl","children":[{"author":"dmix","children":[{"author":"brookst","children":[{"author":"dmix","children":[],"created_at":"2026-04-01T00:09:25.000Z","created_at_i":1775002165,"id":47595149,"options":[],"parent_id":47589119,"points":null,"story_id":47584540,"text":"Capitalism is always underpinned by a strong legal system which is why most criticism is about constraining growth in legislation, not killing off interference outright. Copyright law is a good example of a law that made sense in it's original form but turned into a monster with scope-creep.
Although, if we're being realpolitik, every time government interference grows in scope and corrupts markets, capitalism still gets blamed and people call for more government to fix it (see: housing). So the capitalism vs state capitalism distinction isn't very meaningful in practice.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:45:41.000Z","created_at_i":1774971941,"id":47589119,"options":[],"parent_id":47588577,"points":null,"story_id":47584540,"text":"Where can I see an example of any other kind of capitalism?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:15:33.000Z","created_at_i":1774970133,"id":47588577,"options":[],"parent_id":47587266,"points":null,"story_id":47584540,"text":"Technically state-capitalism since it's an industry created as a result of congress regulating commerce with aggressive IP laws (aka rent-seeking)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:42:01.000Z","created_at_i":1774964521,"id":47587266,"options":[],"parent_id":47587188,"points":null,"story_id":47584540,"text":"That\u2019s capitalism too","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:36:31.000Z","created_at_i":1774964191,"id":47587188,"options":[],"parent_id":47586831,"points":null,"story_id":47584540,"text":"As opposed to the rent-seeking copyright industry where 1% goes to the original creators if you're lucky.","title":null,"type":"comment","url":null},{"author":"satvikpendem","children":[],"created_at":"2026-03-31T13:44:40.000Z","created_at_i":1774964680,"id":47587312,"options":[],"parent_id":47586831,"points":null,"story_id":47584540,"text":"As opposed to what economic system that doesn't do this?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:08:54.000Z","created_at_i":1774962534,"id":47586831,"options":[],"parent_id":47586739,"points":null,"story_id":47584540,"text":"That\u2019s capitalism","title":null,"type":"comment","url":null},{"author":"keybored","children":[],"created_at":"2026-03-31T13:10:00.000Z","created_at_i":1774962600,"id":47586845,"options":[],"parent_id":47586739,"points":null,"story_id":47584540,"text":"The Golden Horde didn\u2019t want opponents to conquer their territory. An irony if you think about it\u2014","title":null,"type":"comment","url":null},{"author":"snapcaster","children":[{"author":"brookst","children":[{"author":"salawat","children":[{"author":"brookst","children":[{"author":"salawat","children":[{"author":"spiderfarmer","children":[{"author":"snapcaster","children":[],"created_at":"2026-04-03T15:08:02.000Z","created_at_i":1775228882,"id":47627522,"options":[],"parent_id":47599163,"points":null,"story_id":47584540,"text":"Would you feel the same way if a professional athlete was saying it? are you just anti-competition in general?","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T10:48:46.000Z","created_at_i":1775040526,"id":47599163,"options":[],"parent_id":47589801,"points":null,"story_id":47584540,"text":"Whenever I see someone on HN preaching about how it's all dog-eat-dog and zero-sum, I imagine them being lonely.
No real friends, no trusted life partner, no kids, no unconditional love. Alone.
Just another soul traveling on an infinite road with lots of signs that point to "happiness," planted there by fellow travelers, never reaching their destination.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:26:03.000Z","created_at_i":1774974363,"id":47589801,"options":[],"parent_id":47589388,"points":null,"story_id":47584540,"text":"Categorical imperative and Golden Rule, or as you may know it from game theory "tit-for-tat" says "hi". The beautiful thing about ethics is that we philosophers intentionally teach it descriptively, but encourage one to choose their own based on context invariance. What this does is create an effective litmus test for detecting shitty people/behavior. You grasping on for dear life to "there's no absolutes" is an act of self-soothing on your own part as you're trying to rationalize your own behavior to provide an ego crumple zone. I, on the other hand, don't intend to leave you that option. That you're having to do it is a Neon sign of your own unethicality in this matter. We get to have nice things when people moderate themselves (we tolerate eventual free access to everything as long as the people who don't want to pay for it don't go and try to replace us economically at scale). When people abuse that, (scrape the Internet, try to sell work product in a way that jeopardizes the environment we create in) the nice thing starts going away, and you've made the world worse.
Welcome to life bucko. Stop being a shitty person and get with the program so we have something to leave behind that has a chance of not making us villains in the eyes of those we eventually leave behind. The trick is doing things the harder way because it's the right way to do it. Not doing it the wrong way because you're pretty sure you can get away with it.
But you're already ethically compromised, so I don't really expect this to do any good except to maybe make the part of you you pointedly ignore start to stir assuming you haven't completely given yourself up to a life of ne'er-do-wellry. Enjoy the enantidromia. Failing that, karma's a bitch.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:01:29.000Z","created_at_i":1774972889,"id":47589388,"options":[],"parent_id":47589333,"points":null,"story_id":47584540,"text":"The funny thing about ethics is there is no absolute, which makes some people uncomfortable. Is it ethical to slice someone with a knife? Does it depend if you're a surgeon or not?
Absolutism + reductionism leads to this kind of nonsense. It is possible that people can disagree about (re)use of culture, including music and print. Therefore it is possible for nuance and context to matter.
Life is a lot easier if you subscribe to a "anyone who disagrees with me on any topic must have no ethics whatsoever and is a BAD person." But it's really not an especially mature worldview.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:58:23.000Z","created_at_i":1774972703,"id":47589333,"options":[],"parent_id":47589102,"points":null,"story_id":47584540,"text":"No, it's ethical people pointing out that if you toss aside ethics for success at all costs, you aren't going to find any sympathy when people start doing the same thing back to you. Live by the sword, die by the sword, as they say.
There is a reason we don't do things. That reason is it makes the world a worse place for everyone. If you are so incredibly out of touch with any semblance of ethics at all; mayhaps you are just a little bit part of the problem.","title":null,"type":"comment","url":null},{"author":"snapcaster","children":[],"created_at":"2026-04-02T15:25:44.000Z","created_at_i":1775143544,"id":47615748,"options":[],"parent_id":47589102,"points":null,"story_id":47584540,"text":"Yeah man I would love to play poker against these people","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:44:57.000Z","created_at_i":1774971897,"id":47589102,"options":[],"parent_id":47587337,"points":null,"story_id":47584540,"text":"Wait, are you saying that it's not hypocritical for my chess opponent to try to protect their king while trying to kill mine? :mind-blown:
Tech people are funny, with these takes that businesses do/should adhere to absolute platonic ideals and follow them blindly regardless of context.","title":null,"type":"comment","url":null},{"author":"vor_","children":[{"author":"snapcaster","children":[],"created_at":"2026-04-01T13:32:28.000Z","created_at_i":1775050348,"id":47600634,"options":[],"parent_id":47593393,"points":null,"story_id":47584540,"text":"No it isn't, have you never played any game or sport? do you fundamentally not understand the concept of competition?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:58:46.000Z","created_at_i":1774990726,"id":47593393,"options":[],"parent_id":47587337,"points":null,"story_id":47584540,"text":"It's definitely still hypocrisy.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:46:05.000Z","created_at_i":1774964765,"id":47587337,"options":[],"parent_id":47586739,"points":null,"story_id":47584540,"text":"No it isn't. It's a competition, making moves that benefit you and attempting to deprive your opponent of the same move is just called competing","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:00:33.000Z","created_at_i":1774962033,"id":47586739,"options":[],"parent_id":47586633,"points":null,"story_id":47584540,"text":"That isn't irony, it's hypocrisy.","title":null,"type":"comment","url":null},{"author":"jaccola","children":[{"author":"brookst","children":[],"created_at":"2026-03-31T15:42:42.000Z","created_at_i":1774971762,"id":47589056,"options":[],"parent_id":47587897,"points":null,"story_id":47584540,"text":"It's really just tech culture like HN that obsesses over solving problems perfectly. From seat belts to DRM to deodorant, most of the world is satisfied with mitigating problems.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:27:52.000Z","created_at_i":1774967272,"id":47587897,"options":[],"parent_id":47586633,"points":null,"story_id":47584540,"text":"I would say not all that ironic. Book publishers, Reddit, Stackoverflow, etc., tried their best to attract customers while not letting others steal their work. Now Anthropic is doing the same.
Unfortunately (for the publishers, at least) it didn't work to stop Anthropic and Anthropic's attempts to prevent others will not work either; there has been much distillation already.
The problem of letting humans read your work but not bots is just impossible to solve perfectly. The more you restrict bots, the more you end up restricting humans, and those humans will go use a competitor when they become pissed off.","title":null,"type":"comment","url":null},{"author":"johnfn","children":[],"created_at":"2026-03-31T14:30:04.000Z","created_at_i":1774967404,"id":47587922,"options":[],"parent_id":47586633,"points":null,"story_id":47584540,"text":"It is absolutely not paranoia. People are distilling Claude code all the time.","title":null,"type":"comment","url":null},{"author":"sheept","children":[{"author":"salawat","children":[],"created_at":"2026-03-31T15:53:16.000Z","created_at_i":1774972396,"id":47589248,"options":[],"parent_id":47588431,"points":null,"story_id":47584540,"text":"Their business shouldn't exist. It was predisposed on non-permissive IP theft. They may have found a judge willing to cop to it not being so, but the rest of the public knows the real score. And most problematically for them, that means the subset of hackerdom that lives by tit-for-tat. One should beware of pissing off gray-hats. Iit's a surefire way to find yourself heading for bad times.","title":null,"type":"comment","url":null},{"author":"gmerc","children":[],"created_at":"2026-03-31T17:17:59.000Z","created_at_i":1774977479,"id":47590569,"options":[],"parent_id":47588431,"points":null,"story_id":47584540,"text":"And conveniently left out they themselves distilled DeepSeek for chinese content into their model....","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:06:05.000Z","created_at_i":1774969565,"id":47588431,"options":[],"parent_id":47586633,"points":null,"story_id":47584540,"text":"It's not really paranoia if it's happening a lot. They wrote a blog post calling several major Chinese AI companies out for distillation.[0] Perhaps it is ironic, but it's within their rights to protect their business, like how they prohibit using Claude Code to make your own Claude Code.[1]
[0]: https://www.anthropic.com/news/detecting-and-preventing-dist...\n[1]: https://news.ycombinator.com/item?id=46578701","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:51:42.000Z","created_at_i":1774961502,"id":47586633,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"Paranoia. And also ironic considering their base LLM is a distillation of the web and books etc etc.","title":null,"type":"comment","url":null},{"author":"crazylogger","children":[],"created_at":"2026-03-31T14:18:39.000Z","created_at_i":1774966719,"id":47587770,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"Why would this be in the client code though?","title":null,"type":"comment","url":null},{"author":"mmaunder","children":[{"author":"cedws","children":[{"author":"mmaunder","children":[{"author":"mmaunder","children":[{"author":"cedws","children":[{"author":"mmaunder","children":[{"author":"BoorishBears","children":[],"created_at":"2026-03-31T23:42:38.000Z","created_at_i":1775000558,"id":47594943,"options":[],"parent_id":47592922,"points":null,"story_id":47584540,"text":"We can be used to refer to people in general, and we know because Anthropic published a post called "Detecting and preventing distillation attacks" a month ago, while calling out 3 AI labs for large scale distillation
https://www.anthropic.com/news/detecting-and-preventing-dist...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:20:13.000Z","created_at_i":1774988413,"id":47592922,"options":[],"parent_id":47592415,"points":null,"story_id":47584540,"text":"Who's we, and how do you know this?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:42:38.000Z","created_at_i":1774986158,"id":47592415,"options":[],"parent_id":47592288,"points":null,"story_id":47584540,"text":"We do know that Anthropic has the ability to detect when their models are being distilled, so there could be some backend mechanism that needs to be tripped to observe certain behaviour. Not possible to confirm though.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:31:50.000Z","created_at_i":1774985510,"id":47592288,"options":[],"parent_id":47592072,"points":null,"story_id":47584540,"text":"Claude Code has a server-side anti-distillation opt-in called fake_tools, but the local code does not show the actual mechanism.
The client sometimes sends anti_distillation: ['fake_tools'] in the request body at services/api/claude.ts:301
The client still sends its normal real tools: allTools at services/api/claude.ts:1711
If the model emits a tool name the client does not actually have, the client turns that into No such tool available errors at services/tools/StreamingToolExecutor.ts:77 and services/tools/toolExecution.ts:369
If Anthropic were literally appending extra normal tool definitions to the live tool set, and Claude used them, that would be user-visible breakage.
That leaves a few more plausible possibilities:
Fake_tools is just the name of the server-side experiment, but the implementation is subtler than \u201cappend fake tools to the real tool list.\u201d
or
The server may inject tool-looking text into hidden prompt context, with separate hidden instructions not to call it.
or
The server may use decoys only in an internal representation that is useful for poisoning traces/training data but not exposed as real executable tools.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:13:13.000Z","created_at_i":1774984393,"id":47592072,"options":[],"parent_id":47588898,"points":null,"story_id":47584540,"text":"Great theory. I'll dig deeper.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:33:50.000Z","created_at_i":1774971230,"id":47588898,"options":[],"parent_id":47588088,"points":null,"story_id":47584540,"text":"Not sure, and not completely convinced of the explanation, but the way this sticks out so obviously makes it look like a honeypot to me.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:43:41.000Z","created_at_i":1774968221,"id":47588088,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"Haven\u2019t looked at the code, but is the server providing the client with a system prompt that it can use, which would contain fake tool definitions when this is enabled? What enables it? And why is the client still functional when it\u2019s giving the server back a system prompt with fake tool definitions? Is the LLM trained to ignore those definitions?
Wonder if they\u2019re also poisoning Sonnet or Opus directly generating simulated agentic conversations.","title":null,"type":"comment","url":null},{"author":"GorbachevyChase","children":[{"author":"cedws","children":[{"author":"lobsterthief","children":[],"created_at":"2026-04-01T12:46:34.000Z","created_at_i":1775047594,"id":47600114,"options":[],"parent_id":47592311,"points":null,"story_id":47584540,"text":"But then the Chinese government has the ultimate say and, propaganda aside, if they don\u2019t like what your product is you might suddenly lose access to said LLM provider.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:33:37.000Z","created_at_i":1774985617,"id":47592311,"options":[],"parent_id":47590931,"points":null,"story_id":47584540,"text":"It made me raise my eyebrows when everyone was rushing to jump to Claude because OpenAI agreed to work with the DoW. Both companies are just as shitty as each other and will resort to underhanded tactics to stay on top.
Go China to be honest. They're the most committed to open AI research and they have more interesting constraints to work under, like restricted access to NVIDIA hardware.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:42:13.000Z","created_at_i":1774978933,"id":47590931,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"I like these guys less every day. The rate limits are so low they are close to not even useful as a provider.","title":null,"type":"comment","url":null},{"author":"3form","children":[],"created_at":"2026-03-31T17:50:54.000Z","created_at_i":1774979454,"id":47591058,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"I was thinking just yesterday that the research that Anthropic was sharing regarding how it's easy to poison training was unlikely to be conducted out of goodness of the heart.","title":null,"type":"comment","url":null},{"author":"jjcm","children":[],"created_at":"2026-03-31T18:37:47.000Z","created_at_i":1774982267,"id":47591619,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"It looks like it worked, fwiw.
The qwen 27b model distilled on Opus 4.6 has some known issues with tool use specifically: https://x.com/KyleHessling1/status/2038695344339611783
Fascinating.","title":null,"type":"comment","url":null},{"author":"torginus","children":[],"created_at":"2026-04-01T13:37:30.000Z","created_at_i":1775050650,"id":47600699,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"This made me think of something - at work, if we wfh, we have to use one of those MITM proxies that intercept HTTPS at the kernel level. Imo such a thing can easily read the traffic and thus is indistinguishable from a distillation attempt from CC's PoV. I've had CC freak out on my machine, and sometimes generate pretty bad results, the CoT is often also not available.
I wonder it CC thinks I'm trying to distill the model. This is a common enough use case that I think the devs at Anthropic should consider.","title":null,"type":"comment","url":null},{"author":"pqtr2","children":[],"created_at":"2026-04-01T18:03:29.000Z","created_at_i":1775066609,"id":47604323,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"Good explanation here https://alex000kim.com/posts/2026-03-31-claude-code-source-l... its a server side flag set by the distillation detection, shouldn't affect most user's sessions. I don't think?","title":null,"type":"comment","url":null},{"author":"huang-b62b5756","children":[],"created_at":"2026-04-02T07:40:34.000Z","created_at_i":1775115634,"id":47611216,"options":[],"parent_id":47585239,"points":null,"story_id":47584540,"text":"It is a RCE, if what\u2019s being injected is an executable and malicious tool","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:28:29.000Z","created_at_i":1774952909,"id":47585239,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"
ANTI_DISTILLATION_CC\n \n This is Anthropic's anti-distillation defence baked into Claude Code. When enabled, it injects anti_distillation: ['fake_tools'] into every API request, which causes the server to silently slip decoy tool definitions into the model's system prompt. The goal: if someone is scraping Claude Code's API traffic to train a competing model, the poisoned training data makes that distillation attempt less useful.Not having to deal with Boris Cherny's UX choices for CC is the cherry on top.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:57:39.000Z","created_at_i":1774961859,"id":47586707,"options":[],"parent_id":47586503,"points":null,"story_id":47584540,"text":"Pi is the best choice for experts and power users, which is not most people.
Claude Code is still the dominant (I didn't say best) agentic harness by a wide margin I think.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:42:12.000Z","created_at_i":1774960932,"id":47586503,"options":[],"parent_id":47585328,"points":null,"story_id":47584540,"text":"I think pi has stolen the top honors, but people consider the Claude code harness very good (at least, better than Cursor)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:39:16.000Z","created_at_i":1774953556,"id":47585328,"options":[],"parent_id":47585251,"points":null,"story_id":47584540,"text":"What's special about Claude Code? Isn't Opus the real magic?
Surely there's nothing here of value compared to the weights except for UX and orchestration?
Couldn't this have just been decompiled anyhow?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:29:38.000Z","created_at_i":1774952978,"id":47585251,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Once the USA wakes up, this will be insane news","title":null,"type":"comment","url":null},{"author":"bkryza","children":[{"author":"raihansaputra","children":[],"created_at":"2026-03-31T10:49:12.000Z","created_at_i":1774954152,"id":47585424,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"i wish that's for their logging/alert. i definitely gauge model's performance by how much those words i type when i'm frustrated in driving claude code.","title":null,"type":"comment","url":null},{"author":"nodja","children":[],"created_at":"2026-03-31T10:56:51.000Z","created_at_i":1774954611,"id":47585505,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"If anyone at anthropic is reading this and wants more logs from me add jfc.","title":null,"type":"comment","url":null},{"author":"BoppreH","children":[{"author":"lou1306","children":[{"author":"noosphr","children":[{"author":"8cvor6j844qw_d6","children":[{"author":"mr_00ff00","children":[],"created_at":"2026-03-31T15:34:11.000Z","created_at_i":1774971251,"id":47588905,"options":[],"parent_id":47585858,"points":null,"story_id":47584540,"text":"My understanding of vibe coding is when someone doesn\u2019t look at the code and just uses prompts until the app \u201clooks and acts\u201d correct.
I doubt you are making regex and not looking at it, even if it was AI generated.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:37:41.000Z","created_at_i":1774957061,"id":47585858,"options":[],"parent_id":47585616,"points":null,"story_id":47584540,"text":"Very likely vibe coded.
I've seen Claude Code went with a regex approach for a similar sentiment-related task.","title":null,"type":"comment","url":null},{"author":"rdiddly","children":[],"created_at":"2026-03-31T14:44:33.000Z","created_at_i":1774968273,"id":47588102,"options":[],"parent_id":47585616,"points":null,"story_id":47584540,"text":"Clbuttic!","title":null,"type":"comment","url":null},{"author":"lou1306","children":[{"author":"noosphr","children":[{"author":"lou1306","children":[],"created_at":"2026-04-01T06:58:44.000Z","created_at_i":1775026724,"id":47597758,"options":[],"parent_id":47592830,"points":null,"story_id":47584540,"text":"> Every conversation is sent to an llm at least a thousand times the size of gpt2 which could one shot this nearly a decade ago.
Yes, but that is _what the product does_. What we are talking about is _telemetry_.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:14:00.000Z","created_at_i":1774988040,"id":47592830,"options":[],"parent_id":47589708,"points":null,"story_id":47584540,"text":">Some people, when confronted with a problem, think \u201cI know, I\u2019ll use regular expressions.\u201d Now they have two problems.
The only time to use a regex is when searching with a human in the loop. All other uses are better handled some other way.
>They could do a full-blown NLP-driven sentiment analysis on every prompt but I reckon it would not be as cost-effective as this.
Every conversation is sent to an llm at least a thousand times the size of gpt2 which could one shot this nearly a decade ago.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:20:17.000Z","created_at_i":1774974017,"id":47589708,"options":[],"parent_id":47585616,"points":null,"story_id":47584540,"text":"> The issue is that you shouldn't be looking for substrings in the first place.
Why? They clearly just want to log conversations that are likely to display extreme user frustration with minimal overhead. They could do a full-blown NLP-driven sentiment analysis on every prompt but I reckon it would not be as cost-effective as this.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:09:28.000Z","created_at_i":1774955368,"id":47585616,"options":[],"parent_id":47585556,"points":null,"story_id":47584540,"text":"The issue isn't that regex are a solution to find a substring. The issue is that you shouldn't be looking for substrings in the first place.
This has buttbuttin energy. Welcome to the 80s I guess.","title":null,"type":"comment","url":null},{"author":"BoppreH","children":[{"author":"hk__2","children":[{"author":"NitpickLawyer","children":[],"created_at":"2026-03-31T14:49:53.000Z","created_at_i":1774968593,"id":47588185,"options":[],"parent_id":47585687,"points":null,"story_id":47584540,"text":"> There\u2019s no point in overengineering it.
I swear this whole thread about regexes is just fake rage at something, and I bet it'd be reversed had they used something heavier (omg, look they're using an LLM call where a simple regex would have worked, lul)...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:16:25.000Z","created_at_i":1774955785,"id":47585687,"options":[],"parent_id":47585626,"points":null,"story_id":47584540,"text":"It\u2019s fast and it matches 80% of the cases. There\u2019s no point in overengineering it.","title":null,"type":"comment","url":null},{"author":"vharuck","children":[{"author":"BoppreH","children":[],"created_at":"2026-03-31T11:21:08.000Z","created_at_i":1774956068,"id":47585737,"options":[],"parent_id":47585713,"points":null,"story_id":47584540,"text":"You're right, I missed the \\b's. Thanks for the correction.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:18:29.000Z","created_at_i":1774955909,"id":47585713,"options":[],"parent_id":47585626,"points":null,"story_id":47584540,"text":"The pattern only matches if both ends are word boundaries. So "diffs" won't match, but "Oh, ffs!" will. It's also why they had to use the pattern "shit(ty|tiest)" instead of just "shit".","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:10:45.000Z","created_at_i":1774955445,"id":47585626,"options":[],"parent_id":47585556,"points":null,"story_id":47584540,"text":"It's fast, but it'll miss a ton of cases. This feels like it would be better served by a prompt instruction, or an additional tiny neural network.
And some of the entries are too short and will create false positives. It'll match the word "offset" ("ffs"), for example. EDIT: no it won't, I missed the \\b. Still sounds weird to me.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:02:01.000Z","created_at_i":1774954921,"id":47585556,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"They're searching for multiple substrings in a single pass, regexes are the optimal solution for that.","title":null,"type":"comment","url":null},{"author":"stingraycharles","children":[{"author":"_fizz_buzz_","children":[{"author":"mycall","children":[{"author":"b112","children":[],"created_at":"2026-03-31T12:15:57.000Z","created_at_i":1774959357,"id":47586215,"options":[],"parent_id":47586148,"points":null,"story_id":47584540,"text":"Did you just complain about bloat, in anything using npm?","title":null,"type":"comment","url":null},{"author":"TeMPOraL","children":[{"author":"MetalSnake","children":[{"author":"ekropotin","children":[{"author":"foldr","children":[{"author":"adamsb6","children":[{"author":"curioussquirrel","children":[],"created_at":"2026-03-31T18:24:04.000Z","created_at_i":1774981444,"id":47591469,"options":[],"parent_id":47587494,"points":null,"story_id":47584540,"text":"Thank you. +1.\nThere are obviously differences and things getting lost or slightly misaligned in the latent space, and these do cause degradation in reasoning quality, but the decline is very small in high resource languages.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:57:51.000Z","created_at_i":1774965471,"id":47587494,"options":[],"parent_id":47587159,"points":null,"story_id":47584540,"text":"They literally just have to subtract the vector for the source language and add the vector for the target.
It\u2019s the original use case for LLMs.","title":null,"type":"comment","url":null},{"author":"ekropotin","children":[{"author":"foldr","children":[{"author":"ekropotin","children":[{"author":"foldr","children":[],"created_at":"2026-03-31T18:35:51.000Z","created_at_i":1774982151,"id":47591595,"options":[],"parent_id":47591493,"points":null,"story_id":47584540,"text":"It\u2019s negligible as far as I can tell. If the LLM can \u201cspeak\u201d the language well then you can prompt it in that language and get more or less the same results as in English.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:26:54.000Z","created_at_i":1774981614,"id":47591493,"options":[],"parent_id":47591176,"points":null,"story_id":47584540,"text":"> it\u2019s clearly not the case that these things only make use of Spanish training data when you prompt them in Spanish.
It\u2019s not! And I\u2019ve never said that.
Anyways, I\u2019m not even sure what we are arguing about, as it\u2019s 100% fact that SOTA models perform better in English, the only interesting question here how much better, is it negligible or actually makes a difference in real world use-cases.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:00:44.000Z","created_at_i":1774980044,"id":47591176,"options":[],"parent_id":47588515,"points":null,"story_id":47584540,"text":"The agent may be translating to English behind the scenes. There are all kinds of hacks layered on top of the underlying models. But however exactly it works, it\u2019s clearly not the case that these things only make use of Spanish training data when you prompt them in Spanish.
Lets take a language with orders of magnitude less training data, like Nahuatl. The following is what happens if I ask Claude Code in Nahuatl to give me an example of a simple program that I can execute inside a web browser (avoiding any direct keyword prompts like 'Javascript' or 'Hello World'). Interestingly, it responds in Spanish. I doubt that it has any relevant training data in Nahuatl. (Apologies if the Nahuatl is wrong \u2013 I just used Google translate.)
\u00bfUelis techmakas se neskayotl tlen se programa tlen amo ouij tlen ueliskia nijchiuas ipan se navegador web?\n\n -----\n\n Entiendo tu pregunta en n\u00e1huatl. Aqu\u00ed un ejemplo sencillo \u2014 una calculadora en HTML puro que puedes abrir directo en el navegador, sin servidor:\n\n <!DOCTYPE html>\n <html>\n <head>\n <meta charset="UTF-8">\n <title>Calculadora</title>\n <style>\n body { font-family: sans-serif; display: flex; justify-content: center; padding: 40px; }\n .calc { border: 1px solid #ccc; padding: 20px; border-radius: 8px; width: 200px; }\n input { width: 100%; margin-bottom: 10px; font-size: 1.2em; text-align: right; }\n button { width: 22%; margin: 2px; padding: 8px; font-size: 1em; cursor: pointer; }\n .wide { width: 47%; }\n </style>\n </head>\n <body>\n <div class="calc">\n <input type="text" id="pantalla" readonly>\n <div>\n <button onclick="tecla('7')">7</button>\n <button onclick="tecla('8')">8</button>\n <button onclick="tecla('9')">9</button>\n <button onclick="tecla('/')">\u00f7</button>\n \n <button onclick="tecla('4')">4</button>\n <button onclick="tecla('5')">5</button>\n <button onclick="tecla('6')">6</button>\n <button onclick="tecla('*')">\u00d7</button>\n \n <button onclick="tecla('1')">1</button>\n <button onclick="tecla('2')">2</button>\n <button onclick="tecla('3')">3</button>\n <button onclick="tecla('-')">\u2212</button>\n \n <button onclick="tecla('0')">0</button>\n <button onclick="tecla('.')">.</button>\n <button onclick="tecla('+')">+</button>\n <button onclick="limpiar()">C</button>\n \n <button class="wide" onclick="calcular()">=</button>\n <button class="wide" onclick="borrar()">\u232b</button>\n </div>\n </div>\n \n <script>\n const pantalla = document.getElementById('pantalla');\n \n function tecla(val) {\n pantalla.value += val;\n }\n \n function calcular() {\n try {\n pantalla.value = eval(pantalla.value);\n } catch {\n pantalla.value = 'Error';\n }\n }\n \n function limpiar() { pantalla.value = ''; }\n function borrar() { pantalla.value = pantalla.value.slice(0, -1); }\n </script>\n </body>\n </html>\n \n Guarda esto como calculadora.html y \u00e1brelo en cualquier navegador \u2014 no necesita servidor ni dependencias. Es un buen punto de partida para aprender HTML,\n CSS y JavaScript.Even if it happens at the edge, on every conversation turn, I may assume non captured small subtleties of meaning over time can accumulate into significant error.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:52:16.000Z","created_at_i":1774983136,"id":47591794,"options":[],"parent_id":47588515,"points":null,"story_id":47584540,"text":"I saw a curious post recently that explored this idea, and showed that it isn\u2019t really the case. The internal layers of the model aren\u2019t really reasoning in English, or in any human language.
Translation in/out of human languages only happens at the edges of the model.
Internal layer activations for the same concept are similar regardless of language, while activations at the top/bottom layers diverge. Meanwhile the pattern is reversed for same language different content.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:11:10.000Z","created_at_i":1774969870,"id":47588515,"options":[],"parent_id":47587159,"points":null,"story_id":47584540,"text":"It\u2019s just a subjective observation.
It just can\u2019t be a case simply because how ML works. In short, the more diverse and high quality texts with reasoning reach examples were in the training set, the better model performs on a given language.
So unless Spanish subset had much more quality-dense examples, to make up for volume, there is no way the quality of reasoning in Spanish is on par with English.
I apologise for the rambling explanation, I sure someone with ML expertise here can it explain it better.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:34:41.000Z","created_at_i":1774964081,"id":47587159,"options":[],"parent_id":47587049,"points":null,"story_id":47584540,"text":"In my experience these models work fine using another language, if it\u2019s a widely spoken one. For example, sometimes I prompt in Spanish, just to practice. It doesn\u2019t seem to\naffect the quality of code generation.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:26:21.000Z","created_at_i":1774963581,"id":47587049,"options":[],"parent_id":47586434,"points":null,"story_id":47584540,"text":"Because 90% of training data was in English and therefore the model perform best in this language.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:35:17.000Z","created_at_i":1774960517,"id":47586434,"options":[],"parent_id":47586249,"points":null,"story_id":47584540,"text":"I talk to it in non-English. But have rules to have everything in code and documentation in english. Only speaking with me should use my native language. Why would that be a problem?","title":null,"type":"comment","url":null},{"author":"formerly_proven","children":[],"created_at":"2026-03-31T12:55:29.000Z","created_at_i":1774961729,"id":47586678,"options":[],"parent_id":47586249,"points":null,"story_id":47584540,"text":"In my experience agents tend to (counterintuitively) perform better when the business language is not English / does not match the code's language. I'm assuming the increased attention mitigates the higher "cognitive" load.","title":null,"type":"comment","url":null},{"author":"cryptonector","children":[],"created_at":"2026-03-31T15:42:45.000Z","created_at_i":1774971765,"id":47589059,"options":[],"parent_id":47586249,"points":null,"story_id":47584540,"text":"Claude handles human languages other than English just fine.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:19:18.000Z","created_at_i":1774959558,"id":47586249,"options":[],"parent_id":47586148,"points":null,"story_id":47584540,"text":"We're talking about Claude Code. If you're coding and not writing or thinking in English, the agents and people reading that code will have bigger problems than a regexp missing a swear word :).","title":null,"type":"comment","url":null},{"author":"doublesocket","children":[{"author":"cogman10","children":[{"author":"mycall","children":[],"created_at":"2026-04-01T13:22:12.000Z","created_at_i":1775049732,"id":47600511,"options":[],"parent_id":47587771,"points":null,"story_id":47584540,"text":"Exactly this. Unicode is a big beast to consider in regex concats.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:18:41.000Z","created_at_i":1774966721,"id":47587771,"options":[],"parent_id":47586629,"points":null,"story_id":47584540,"text":"I think it will depend on the language. There are a few non-latin languages where a simple word search likely won't be enough for a regex to properly apply.","title":null,"type":"comment","url":null},{"author":"astrocat","children":[{"author":"benlivengood","children":[],"created_at":"2026-03-31T15:09:42.000Z","created_at_i":1774969782,"id":47588498,"options":[],"parent_id":47587914,"points":null,"story_id":47584540,"text":"You can literally | together every street address or other string you want to match in a giant disjunction, and then run a DFA/NFA minimization over that to get it down to a reasonable size. Maybe there are some fast regex simplification algorithms as well, but working directly with the finite automata has decades of research and probably can be more fully optimized.","title":null,"type":"comment","url":null},{"author":"doublesocket","children":[],"created_at":"2026-03-31T18:22:32.000Z","created_at_i":1774981352,"id":47591455,"options":[],"parent_id":47587914,"points":null,"story_id":47584540,"text":"This was many moons ago, written in perl. From memory we used Regexp::Trie -\nhttps://metacpan.org/release/DANKOGAI/Regexp-Trie-0.02/view/...
We used it to tokenize search input and combined it with a solr backend. Worked really remarkably well.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:29:05.000Z","created_at_i":1774967345,"id":47587914,"options":[],"parent_id":47586629,"points":null,"story_id":47584540,"text":"woah. This is a regex use I've never heard of. I'd absolutely love to see a writeup on this approach - how its done and when it's useful.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:51:29.000Z","created_at_i":1774961489,"id":47586629,"options":[],"parent_id":47586148,"points":null,"story_id":47584540,"text":"Supporting 10 different languages in regex is a drop in the ocean. The regex can be generated programmatically and you can compress regexes easily. We used to have a compressed regex that could match any placename or street name in the UK in a few MB of RAM. It was silly quick.","title":null,"type":"comment","url":null},{"author":"crimsonnoodle58","children":[],"created_at":"2026-03-31T13:06:45.000Z","created_at_i":1774962405,"id":47586805,"options":[],"parent_id":47586148,"points":null,"story_id":47584540,"text":"They only need to look at one language to get a statistically meaningful picture into common flaws with their model(s) or application.
If they want to drill down to flaws that only affect a particular language, then they could add a regex for that as well/instead.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:08:29.000Z","created_at_i":1774958909,"id":47586148,"options":[],"parent_id":47585952,"points":null,"story_id":47584540,"text":"The problem with regex is multi-language support and how big the regex will bloat if you to support even 10 languages.","title":null,"type":"comment","url":null},{"author":"crem","children":[],"created_at":"2026-03-31T14:53:28.000Z","created_at_i":1774968808,"id":47588230,"options":[],"parent_id":47585952,"points":null,"story_id":47584540,"text":"More likely their LLM Agent just produced that regex and they didn't even notice.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:47:19.000Z","created_at_i":1774957639,"id":47585952,"options":[],"parent_id":47585589,"points":null,"story_id":47584540,"text":"They probably have statistics on it and saw that certain phrases happen over and over so why waste compute on inference.","title":null,"type":"comment","url":null},{"author":"Foobar8568","children":[{"author":"matkoniecz","children":[],"created_at":"2026-03-31T12:11:20.000Z","created_at_i":1774959080,"id":47586172,"options":[],"parent_id":47585956,"points":null,"story_id":47584540,"text":"> a few seconds won't be a big deal
it is not that slow","title":null,"type":"comment","url":null},{"author":"plorntus","children":[],"created_at":"2026-03-31T12:59:50.000Z","created_at_i":1774961990,"id":47586728,"options":[],"parent_id":47585956,"points":null,"story_id":47584540,"text":"Depends what its used by, if I recall theres an `/insights` command/skill built in whatever you want to call it that generates a HTML file. I believe it gives you stats on when you're frustrated with it and (useless) suggestions on how to "use claude better".
Additionally after looking at the source it looks like a lot of Anthropics own internal test tooling/debug (ie. stuff stripped out at build time) is in this source mapping. Theres one part that prompts their own users (or whatever) to use a report issue command whenever frustration is detected. It's possible its using it for this.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:47:49.000Z","created_at_i":1774957669,"id":47585956,"options":[],"parent_id":47585589,"points":null,"story_id":47584540,"text":"Why do you need to do it at the client side? You are leaking so much information on the client side. \nAnd considering the speed of Claude code, if you really want to do on the client side, a few seconds won't be a big deal.","title":null,"type":"comment","url":null},{"author":"orphea","children":[{"author":"jflynn2","children":[{"author":"orphea","children":[{"author":"gf000","children":[],"created_at":"2026-03-31T13:17:14.000Z","created_at_i":1774963034,"id":47586932,"options":[],"parent_id":47586684,"points":null,"story_id":47584540,"text":"I think it's a very reasonable tradeoff, getting 99% of true positives at the fraction of cost (both runtime and engineering).
Besides, they probably do a separate analysis on server side either way, so they can check a true positive to false positive ratio.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:55:52.000Z","created_at_i":1774961752,"id":47586684,"options":[],"parent_id":47586222,"points":null,"story_id":47584540,"text":"This is assuming the regex is doing a good job. It is not. Also you can embed a very tiny model if you really want to flag as many negatives as possible (I don't know anthropic's goal with this) - it would be quick and free.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:16:46.000Z","created_at_i":1774959406,"id":47586222,"options":[],"parent_id":47586073,"points":null,"story_id":47584540,"text":"Better question - why would you call an LLM (expensive in compute terms) for something that a regex can do (cheap in compute terms)
Regex is going to be something like 10,000 times quicker than the quickest LLM call, multiply that by billions of prompts","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:58:31.000Z","created_at_i":1774958311,"id":47586073,"options":[],"parent_id":47585589,"points":null,"story_id":47584540,"text":"It looks like it's just for logging, why does it need to block?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:05:59.000Z","created_at_i":1774955159,"id":47585589,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Because they want it to be executed quickly and cheaply without blocking the workflow? Doesn\u2019t seem very weird to me at all.","title":null,"type":"comment","url":null},{"author":"codegladiator","children":[{"author":"argee","children":[{"author":"eadler","children":[{"author":"freedomben","children":[{"author":"SmellTheGlove","children":[{"author":"eszed","children":[],"created_at":"2026-03-31T15:43:45.000Z","created_at_i":1774971825,"id":47589083,"options":[],"parent_id":47588417,"points":null,"story_id":47584540,"text":"Ha! Where I'm from a "dolly" was the two-wheeled thing. The four-wheeler thing wasn't common before big-boxes took over the hardware business, but I think my dad would have called it a "cart", maybe a "hand-cart".","title":null,"type":"comment","url":null},{"author":"sethammons","children":[],"created_at":"2026-04-01T09:16:12.000Z","created_at_i":1775034972,"id":47598604,"options":[],"parent_id":47588417,"points":null,"story_id":47584540,"text":"Grew up with two wheeled: dolly and four wheeled: piano dolly. Was an adult when I heard hand-truck. I prefer dolly. Nicer mouth feel.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:05:09.000Z","created_at_i":1774969509,"id":47588417,"options":[],"parent_id":47586270,"points":null,"story_id":47584540,"text":"Isn\u2019t a dolly a flat 4 wheeled platform thingy? A hand truck is the two wheeled thing that tilts back.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:21:23.000Z","created_at_i":1774959683,"id":47586270,"options":[],"parent_id":47585926,"points":null,"story_id":47584540,"text":"Depending on the region you live in, it's also frequently called a "dolly"","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:44:43.000Z","created_at_i":1774957483,"id":47585926,"options":[],"parent_id":47585767,"points":null,"story_id":47584540,"text":"I never knew the name of that device.
Thanks","title":null,"type":"comment","url":null},{"author":"istoleabread","children":[{"author":"svnt","children":[{"author":"apgwoz","children":[],"created_at":"2026-03-31T15:02:47.000Z","created_at_i":1774969367,"id":47588386,"options":[],"parent_id":47588248,"points":null,"story_id":47584540,"text":"It\u2019s also deterministic, unlike llms\u2026","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:54:35.000Z","created_at_i":1774968875,"id":47588248,"options":[],"parent_id":47585961,"points":null,"story_id":47584540,"text":"Yeah it\u2019s called a regex. With a lot of human assistance it can do less but fits in smaller spaces and doesn\u2019t break down.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:48:04.000Z","created_at_i":1774957684,"id":47585961,"options":[],"parent_id":47585767,"points":null,"story_id":47584540,"text":"Do we have a hand llm perchance?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:25:01.000Z","created_at_i":1774956301,"id":47585767,"options":[],"parent_id":47585677,"points":null,"story_id":47584540,"text":"That\u2019s what they do. Ever heard of a hand truck?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:15:02.000Z","created_at_i":1774955702,"id":47585677,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"what you are suggesting would be like a truck company using trucks to move things within the truck","title":null,"type":"comment","url":null},{"author":"draxil","children":[],"created_at":"2026-03-31T11:17:51.000Z","created_at_i":1774955871,"id":47585709,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Good to have more than a hammer in your toolbox!","title":null,"type":"comment","url":null},{"author":"ojr","children":[],"created_at":"2026-03-31T11:46:36.000Z","created_at_i":1774957596,"id":47585946,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"I used regexes in a similar way but my implementation was vibecoded, hmmm, using your analysis Claude Code writes code by hand.","title":null,"type":"comment","url":null},{"author":"blks","children":[{"author":"orphea","children":[{"author":"vntok","children":[],"created_at":"2026-03-31T13:33:44.000Z","created_at_i":1774964024,"id":47587146,"options":[],"parent_id":47586098,"points":null,"story_id":47584540,"text":"They evidently ran a statistical analysis and determined that virtually no one uses those phrases as a quick retort to a model's unsatisfying answer... so they don't need to optimize for them.","title":null,"type":"comment","url":null},{"author":"arcfour","children":[],"created_at":"2026-03-31T13:42:51.000Z","created_at_i":1774964571,"id":47587286,"options":[],"parent_id":47586098,"points":null,"story_id":47584540,"text":"It is exceedingly obvious that the goal here is to catch at least 75-80% of negative sentiment and not to be exhaustive and pedantic and think of every possible way someone could express themselves.","title":null,"type":"comment","url":null},{"author":"Zamaamiro","children":[{"author":"orphea","children":[{"author":"zwirbl","children":[],"created_at":"2026-03-31T14:56:02.000Z","created_at_i":1774968962,"id":47588261,"options":[],"parent_id":47587785,"points":null,"story_id":47584540,"text":"It compares to lowercase input, so doesn't matter. The rest is still valid","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:19:55.000Z","created_at_i":1774966795,"id":47587785,"options":[],"parent_id":47587453,"points":null,"story_id":47584540,"text":"75-80% [1], 90%, 99% [2]. In other words, no one has any idea.
I doubt it's anywhere that high because even if you don't write anything fancy and simply capitalize the first word like you'd normally do at the beginning of a sentence, the regex won't flag it.
Anyway, I don't really care, might just as well be 99.99%. This is not a hill I'm going to die on :P
[1]: https://news.ycombinator.com/item?id=47587286
[2]: https://news.ycombinator.com/item?id=47586932","title":null,"type":"comment","url":null},{"author":"morkalork","children":[],"created_at":"2026-03-31T15:23:22.000Z","created_at_i":1774970602,"id":47588706,"options":[],"parent_id":47587453,"points":null,"story_id":47584540,"text":"Except that it's a list of English keywords. Swearing at the computer is the one thing I'll hear devs switch back to their native language for constantly","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:55:25.000Z","created_at_i":1774965325,"id":47587453,"options":[],"parent_id":47586098,"points":null,"story_id":47584540,"text":"Classic over-engineering. Their approach is just fine 90% of the time for the use case it\u2019s intended for.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:02:31.000Z","created_at_i":1774958551,"id":47586098,"options":[],"parent_id":47586049,"points":null,"story_id":47584540,"text":"Then they made it wrong. For example, "What the actual fuck?" is not getting flagged, neither is "What the *fuck*".","title":null,"type":"comment","url":null},{"author":"mohsen1","children":[],"created_at":"2026-03-31T13:55:14.000Z","created_at_i":1774965314,"id":47587447,"options":[],"parent_id":47586049,"points":null,"story_id":47584540,"text":"Maybe hard to believe but not everyone is speaking English to Claude","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:56:08.000Z","created_at_i":1774958168,"id":47586049,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Because they actually want it to work 100% of the time and cost nothing.","title":null,"type":"comment","url":null},{"author":"sumtechguy","children":[],"created_at":"2026-03-31T11:56:42.000Z","created_at_i":1774958202,"id":47586057,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"hmm not a terrible idea (I think).
You have a semi expensive process. But you want to keep particular known context out. So a quick and dirty search just in front of the expensive process. So instead of 'figure sentiment (20seconds)'. You have 'quick check sentiment (<1sec)' then do the 'figure sentiment v2 (5seconds)'. Now if it is just pure regex then your analogy would hold up just fine.
I could see me totally making a design choice like that.","title":null,"type":"comment","url":null},{"author":"pfortuny","children":[],"created_at":"2026-03-31T12:22:11.000Z","created_at_i":1774959731,"id":47586285,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"They had the problem of sentiment analysis. They use regexes.
You know the drill.","title":null,"type":"comment","url":null},{"author":"throwaw12","children":[],"created_at":"2026-03-31T12:27:56.000Z","created_at_i":1774960076,"id":47586342,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"because impact of WTF might be lost in the result of the analysis if you solely rely on LLM.
parsing WTF with regex also signifies the impact and reduces the noise in metrics
"determinism > non-determinism" when you are analysing the sentiment, why not make some things more deterministic.
Cool thing about this solution, is that you can evaluate LLM sentiment accuracy against regex based approach and analyse discrepancies","title":null,"type":"comment","url":null},{"author":"kjshsh123","children":[],"created_at":"2026-03-31T12:30:37.000Z","created_at_i":1774960237,"id":47586377,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Using regex with LLMs isn't uncommon at all.","title":null,"type":"comment","url":null},{"author":"floralhangnail","children":[{"author":"geon","children":[{"author":"mmh0000","children":[],"created_at":"2026-03-31T15:51:26.000Z","created_at_i":1774972286,"id":47589218,"options":[],"parent_id":47588712,"points":null,"story_id":47584540,"text":"The Clbuttical problem[1]
[1] https://en.wikipedia.org/wiki/Scunthorpe_problem","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:23:28.000Z","created_at_i":1774970608,"id":47588712,"options":[],"parent_id":47586761,"points":null,"story_id":47584540,"text":"buttbuttination","title":null,"type":"comment","url":null},{"author":"raw_anon_1111","children":[],"created_at":"2026-03-31T15:51:35.000Z","created_at_i":1774972295,"id":47589222,"options":[],"parent_id":47586761,"points":null,"story_id":47584540,"text":"I just went to expertSexChange.com\u2026","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:02:27.000Z","created_at_i":1774962147,"id":47586761,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Well, regex doesn't hallucinate....right?","title":null,"type":"comment","url":null},{"author":"mghackerlady","children":[],"created_at":"2026-03-31T13:08:46.000Z","created_at_i":1774962526,"id":47586829,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"More like a car company transporting their shipments by truck. It's more efficient","title":null,"type":"comment","url":null},{"author":"__alexs","children":[],"created_at":"2026-03-31T13:42:43.000Z","created_at_i":1774964563,"id":47587281,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Using some ML to derive a sentiment regex seems like a good actually?","title":null,"type":"comment","url":null},{"author":"lazysheepherd","children":[],"created_at":"2026-03-31T13:50:40.000Z","created_at_i":1774965040,"id":47587396,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Because they are engineers?\nThe difference between an engineer and a hobbyist is an engineer has to optimize the cost.
As they say: any idiot can build a bridge that stands, only an engineer can build a bridge that barely stands.","title":null,"type":"comment","url":null},{"author":"feketegy","children":[],"created_at":"2026-03-31T14:05:20.000Z","created_at_i":1774965920,"id":47587587,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"It's all regex anyways","title":null,"type":"comment","url":null},{"author":"harikb","children":[],"created_at":"2026-03-31T14:10:20.000Z","created_at_i":1774966220,"id":47587652,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Not everything done by claude-code is decided by LLM. They need the wrapper to be deterministic (or one-time generated) code?","title":null,"type":"comment","url":null},{"author":"irthomasthomas","children":[{"author":"irthomasthomas","children":[],"created_at":"2026-03-31T17:27:00.000Z","created_at_i":1774978020,"id":47590695,"options":[],"parent_id":47587699,"points":null,"story_id":47584540,"text":"Actually, this could be a case where its useful. Even it only catches half the complaints, that's still a lot of data, far more than ordinary telemetry used to collect.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:13:12.000Z","created_at_i":1774966392,"id":47587699,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"This just proves its vibe coded because LLMs love writing solutions like that. I probably have a hundred examples just like it in my history.","title":null,"type":"comment","url":null},{"author":"j45","children":[],"created_at":"2026-03-31T14:13:17.000Z","created_at_i":1774966397,"id":47587702,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Asking a non deterministic software to act like a deterministic one (regex) can be a significantly higher use of tokens/compute for no benefit.
Some things will be much better with inference, others won\u2019t be.","title":null,"type":"comment","url":null},{"author":"intended","children":[],"created_at":"2026-03-31T14:15:05.000Z","created_at_i":1774966505,"id":47587733,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"The amount of trust and safety work that depends on google translate and the humble regex, beggars the imagination.","title":null,"type":"comment","url":null},{"author":"lopsotronic","children":[{"author":"cyanydeez","children":[{"author":"true_religion","children":[{"author":"sebastiennight","children":[],"created_at":"2026-03-31T15:19:03.000Z","created_at_i":1774970343,"id":47588640,"options":[],"parent_id":47588035,"points":null,"story_id":47584540,"text":"En toute honn\u00eatet\u00e9, je pense avoir dit "damn it" plus d'une fois \u00e0 chat g\u00e9p\u00e9t\u00e9 avant de fermer la fen\u00eatre dans un acc\u00e8s de rage","title":null,"type":"comment","url":null},{"author":"ben_w","children":[{"author":"sebastiennight","children":[{"author":"bandie91","children":[],"created_at":"2026-03-31T19:18:57.000Z","created_at_i":1774984737,"id":47592139,"options":[],"parent_id":47591168,"points":null,"story_id":47584540,"text":"dark chocolate is "\u00e9tcsokol\u00e1d\u00e9" literally edible-chocolate in Hungarian.
i heared the throat-cleaning "Negr\u00f3" candy (marketed by a chimney sweeper man with soot-covered face) was usually which hurt English-speaking people's self-deprecating sensitivities.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:59:57.000Z","created_at_i":1774979997,"id":47591168,"options":[],"parent_id":47588973,"points":null,"story_id":47584540,"text":"I just went through quite an adventure trying to translate back and forth from/to Hungarian to/from different languages to figure out which Hungarian word you meant, and arrived at the conclusion that this language is encrypted against human comprehension.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:37:54.000Z","created_at_i":1774971474,"id":47588973,"options":[],"parent_id":47588035,"points":null,"story_id":47584540,"text":"The up-side of the US market is (almost) everyone there speaks English. The down side is, that includes all the well-networked pearl-clutchers. Europe (including France) will have the same people, but it's harder to coordinate a network of pearl-clutching between some saying "Il faut prot\u00e9ger nos enfants de cette vulgarit\u00e9!" and others saying "\u0397 \u03c4\u03b7\u03bb\u03b5\u03cc\u03c1\u03b1\u03c3\u03b7 \u03ba\u03b1\u03b9 \u03c4\u03b1 \u03bc\u03ad\u03c3\u03b1 \u03b5\u03bd\u03b7\u03bc\u03ad\u03c1\u03c9\u03c3\u03b7\u03c2 \u03b4\u03b9\u03b1\u03c3\u03c4\u03c1\u03b5\u03b2\u03bb\u03ce\u03bd\u03bf\u03c5\u03bd \u03c4\u03b9\u03c2 \u03b1\u03be\u03af\u03b5\u03c2 \u03bc\u03b1\u03c2!" even when they care about the exact same media.
For headlines, that's enough.
For what's behind the pearl-clutching, for what leads to the headlines pandering to them being worth writing, I agree with everyone else on this thread saying a simple word list is weird and probably pointless. Not just for false-negatives, but also false-positives: the Latin influence on many European languages leads to one very big politically-incorrect-in-the-USA problem for all the EU products talking about anything "black" (which includes what's printed on some brands of dark chocolate, one of which I saw in Hungary even though Hungarian isn't a Latin language but an Ugric language and only takes influences from Latin).","title":null,"type":"comment","url":null},{"author":"bigbuppo","children":[],"created_at":"2026-03-31T16:31:46.000Z","created_at_i":1774974706,"id":47589883,"options":[],"parent_id":47588035,"points":null,"story_id":47584540,"text":"There are only Americans on the internet.","title":null,"type":"comment","url":null},{"author":"tomaskafka","children":[{"author":"z500","children":[],"created_at":"2026-03-31T18:54:13.000Z","created_at_i":1774983253,"id":47591820,"options":[],"parent_id":47591726,"points":null,"story_id":47584540,"text":"It's like wiping your arse with silk.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:46:09.000Z","created_at_i":1774982769,"id":47591726,"options":[],"parent_id":47588035,"points":null,"story_id":47584540,"text":"Nom de dieu de putain de bordel de merde de saloperie de connard d'encul\u00e9 de ta m\u00e8re.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:39:32.000Z","created_at_i":1774967972,"id":47588035,"options":[],"parent_id":47587969,"points":null,"story_id":47584540,"text":"This door lock doesn\u2019t even work against people speaking French, so I think they could have tried a mite harder.","title":null,"type":"comment","url":null},{"author":"themafia","children":[{"author":"jacquesm","children":[],"created_at":"2026-03-31T20:31:16.000Z","created_at_i":1774989076,"id":47593061,"options":[],"parent_id":47592037,"points":null,"story_id":47584540,"text":"That's like saying you can use a chisel for woodworking.","title":null,"type":"comment","url":null},{"author":"senderista","children":[{"author":"themafia","children":[],"created_at":"2026-04-01T06:13:49.000Z","created_at_i":1775024029,"id":47597440,"options":[],"parent_id":47597190,"points":null,"story_id":47584540,"text":"And Claude can't tell at all.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T05:34:32.000Z","created_at_i":1775021672,"id":47597190,"options":[],"parent_id":47592037,"points":null,"story_id":47584540,"text":"Only a native speaker can tell if you swear well in a foreign language.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:10:38.000Z","created_at_i":1774984238,"id":47592037,"options":[],"parent_id":47587969,"points":null,"story_id":47584540,"text":"Yea.. but.. in English only.
Fortunately I can swear pretty well in Spanish.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:33:48.000Z","created_at_i":1774967628,"id":47587969,"options":[],"parent_id":47587915,"points":null,"story_id":47584540,"text":"So, think of it as a business man: You don't really care if your customers swear or whatever, but you know that it'll generate bad headlines. So you gotta do something. Just like a door lock isn't designed for a master criminal, you don't need to design your filter for some master swearer; no, you design it good enough that it gives the impression that further tries are futile.
So yeah, you do what's less intesive to the cpu, but also, you do what's enough to prevent the majority of the concerns where a screenshot or log ends up showing blatant "unmoral" behavior.","title":null,"type":"comment","url":null},{"author":"mlmonkey","children":[{"author":"jonbwhite","children":[{"author":"mlmonkey","children":[],"created_at":"2026-03-31T16:57:13.000Z","created_at_i":1774976233,"id":47590257,"options":[],"parent_id":47589565,"points":null,"story_id":47584540,"text":"I guess I need reading glasses ... :-D","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:12:27.000Z","created_at_i":1774973547,"id":47589565,"options":[],"parent_id":47589403,"points":null,"story_id":47584540,"text":"OP is saying that in their experience it is more like eight orders of magnitude","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:02:43.000Z","created_at_i":1774972963,"id":47589403,"options":[],"parent_id":47587915,"points":null,"story_id":47584540,"text":"> Someone said 10,000x slower, but that's off - in my experience - by about four orders of magnitude.
You do know that 10,000x _is_ four orders of magnitude, right? :-D","title":null,"type":"comment","url":null},{"author":"noprof6691","children":[{"author":"FuckButtons","children":[{"author":"noprof6691","children":[],"created_at":"2026-04-01T08:15:31.000Z","created_at_i":1775031331,"id":47598239,"options":[],"parent_id":47589839,"points":null,"story_id":47584540,"text":"BUT THEY'RE ALREADY RUNNING IT THROUGH THE LLM.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:28:24.000Z","created_at_i":1774974504,"id":47589839,"options":[],"parent_id":47589746,"points":null,"story_id":47584540,"text":"because a regex on the client is free vs gpu compute is absolutely not.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:21:56.000Z","created_at_i":1774974116,"id":47589746,"options":[],"parent_id":47587915,"points":null,"story_id":47584540,"text":"They're sending it to an llm anyway tho? Not sure why they wouldn't just add a sentiment field to the requested response shape.","title":null,"type":"comment","url":null},{"author":"wcrossbow","children":[],"created_at":"2026-03-31T16:46:56.000Z","created_at_i":1774975616,"id":47590109,"options":[],"parent_id":47587915,"points":null,"story_id":47584540,"text":"If it\u2019s good enough it\u2019s good enough, but just like there are many more options than going full blown LLM or just use a regex there are more options than transpile a massive Python stack to TS or give up.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:29:16.000Z","created_at_i":1774967356,"id":47587915,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"The difference in response time - especially versus a regex running locally - is really difficult to express to someone who hasn't made much use of LLM calls in their natural language projects.
Someone said 10,000x slower, but that's off - in my experience - by about four orders of magnitude. And that's average, it gets much worse.
Now personally I would have maybe made a call through a "traditional" ML widget (scikit, numpy, spaCy, fastText, sentence-transformer, etc) but - for me anyway - that whole entire stack is Python. Transpiling all that to TS might be a maintenance burden I don't particularly feel like taking on. And on client facing code I'm not really sure it's even possible.","title":null,"type":"comment","url":null},{"author":"sfn42","children":[],"created_at":"2026-03-31T14:37:33.000Z","created_at_i":1774967853,"id":47588007,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"It's almost as if LLMs are unreliable","title":null,"type":"comment","url":null},{"author":"apgwoz","children":[],"created_at":"2026-03-31T15:01:36.000Z","created_at_i":1774969296,"id":47588351,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"> That's like a truck company using horses to transport parts. Weird choice.
Easy way to claim more \u201chorse power.\u201d","title":null,"type":"comment","url":null},{"author":"makeitrain","children":[],"created_at":"2026-03-31T15:12:57.000Z","created_at_i":1774969977,"id":47588546,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Don\u2019t worry, they used an llm to generate the regex.","title":null,"type":"comment","url":null},{"author":"make3","children":[],"created_at":"2026-03-31T15:28:05.000Z","created_at_i":1774970885,"id":47588800,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"it's like a faster than light spaceship company using horses. There's been infinite solutions to do this better even CPU only for years lol.","title":null,"type":"comment","url":null},{"author":"nojs","children":[{"author":"cryptonector","children":[],"created_at":"2026-03-31T15:45:41.000Z","created_at_i":1774971941,"id":47589118,"options":[],"parent_id":47588970,"points":null,"story_id":47584540,"text":"Wow, that's horrible.","title":null,"type":"comment","url":null},{"author":"lanbin","children":[{"author":"foodevl","children":[],"created_at":"2026-03-31T18:17:43.000Z","created_at_i":1774981063,"id":47591392,"options":[],"parent_id":47589429,"points":null,"story_id":47584540,"text":"It is not a tricky problem because it has a simple and obvious solution: do not filter or block usage just because the input includes a word like "gun".","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:04:02.000Z","created_at_i":1774973042,"id":47589429,"options":[],"parent_id":47588970,"points":null,"story_id":47584540,"text":"This is a tricky problem, I mean, Pinyin also uses the English alphabet.","title":null,"type":"comment","url":null},{"author":"toraway","children":[],"created_at":"2026-03-31T21:26:34.000Z","created_at_i":1774992394,"id":47593675,"options":[],"parent_id":47588970,"points":null,"story_id":47584540,"text":"... and closed for inactivity like basically every issue in the repo, of course.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:37:37.000Z","created_at_i":1774971457,"id":47588970,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Oh it\u2019s worse than that. This one ended up getting my account banned: https://github.com/anthropics/claude-code/issues/22284","title":null,"type":"comment","url":null},{"author":"raw_anon_1111","children":[],"created_at":"2026-03-31T15:48:58.000Z","created_at_i":1774972138,"id":47589178,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Cloud hosted call centers using LLMs is one of my specialties. While I use an LLM for more nuanced sentiment analysis, I definitely use a list of keywords as a first level filter.","title":null,"type":"comment","url":null},{"author":"pdntspa","children":[],"created_at":"2026-03-31T15:50:45.000Z","created_at_i":1774972245,"id":47589204,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"LLMs cost money, regular expressions are free. It really isn't so strange.","title":null,"type":"comment","url":null},{"author":"nitekode","children":[],"created_at":"2026-03-31T16:22:59.000Z","created_at_i":1774974179,"id":47589752,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"A lot if things dont make sense until you involve scale. Regex could be good enough do give a general gist.","title":null,"type":"comment","url":null},{"author":"ldobre","children":[],"created_at":"2026-03-31T16:33:00.000Z","created_at_i":1774974780,"id":47589905,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"It's more like a truck company using people to transport some parts. \nI could be wrong here, but I bet this happens in Volvo's fabrics a lot.","title":null,"type":"comment","url":null},{"author":"arnarbi","children":[],"created_at":"2026-03-31T18:17:50.000Z","created_at_i":1774981070,"id":47591395,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"It's more like workers on a large oil tanker using bicycles to move around it, rather than trying to use another oil tanker.","title":null,"type":"comment","url":null},{"author":"scotty79","children":[],"created_at":"2026-03-31T19:01:27.000Z","created_at_i":1774983687,"id":47591927,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"As far as I can tell they do nothing with it. They just log it.","title":null,"type":"comment","url":null},{"author":"artrockalter","children":[],"created_at":"2026-03-31T20:39:01.000Z","created_at_i":1774989541,"id":47593181,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"LLMs are good at writing complex regex, from my experience","title":null,"type":"comment","url":null},{"author":"slashdave","children":[],"created_at":"2026-04-01T05:41:57.000Z","created_at_i":1775022117,"id":47597226,"options":[],"parent_id":47585535,"points":null,"story_id":47584540,"text":"Maybe. Could just be a pre filter.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:59:49.000Z","created_at_i":1774954789,"id":47585535,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"An LLM company using regexes for sentiment analysis? That's like a truck company using horses to transport parts. Weird choice.","title":null,"type":"comment","url":null},{"author":"sreekanth850","children":[],"created_at":"2026-03-31T11:06:13.000Z","created_at_i":1774955173,"id":47585591,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Glad abusing words in my list are not in that. but its surprising that they use regex for sentiments.","title":null,"type":"comment","url":null},{"author":"moontear","children":[{"author":"conception","children":[],"created_at":"2026-03-31T12:35:17.000Z","created_at_i":1774960517,"id":47586435,"options":[],"parent_id":47585650,"points":null,"story_id":47584540,"text":"/feedback works for that i believe","title":null,"type":"comment","url":null},{"author":"zx8080","children":[{"author":"Paradigma11","children":[{"author":"jollymonATX","children":[],"created_at":"2026-03-31T15:30:20.000Z","created_at_i":1774971020,"id":47588836,"options":[],"parent_id":47587393,"points":null,"story_id":47584540,"text":"This leak just contributed to a new former customer, me. Flagging these phrases may explain exactly why I noticed cc almost immediatly change into grok lvl shit and never recover. Seriously wtf. (flagged again lol)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:50:30.000Z","created_at_i":1774965030,"id":47587393,"options":[],"parent_id":47586984,"points":null,"story_id":47584540,"text":"It also increases the number of former customers.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:21:35.000Z","created_at_i":1774963295,"id":47586984,"options":[],"parent_id":47585650,"points":null,"story_id":47584540,"text":"WTF per minute strongly correlates to an increased token spending.
It may be decided at Anthropic at some moment to increase wtf/min metric, not decrease.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:12:45.000Z","created_at_i":1774955565,"id":47585650,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"I don't know about avoided, this kind of represents the WTF per minute code quality measurement. When I write WTF as a response to Claude, I would actually love if an Antrhopic engineer would take a look at what mess Claude has created.","title":null,"type":"comment","url":null},{"author":"dheerajmp","children":[],"created_at":"2026-03-31T11:28:08.000Z","created_at_i":1774956488,"id":47585794,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Yeah, this is crazy","title":null,"type":"comment","url":null},{"author":"samuelknight","children":[{"author":"dijit","children":[],"created_at":"2026-03-31T12:03:13.000Z","created_at_i":1774958593,"id":47586105,"options":[],"parent_id":47585954,"points":null,"story_id":47584540,"text":"It's actually pretty common for old sysadmin code too..
You could always tell when a sysadmin started hacking up some software by the if-else nesting chains.","title":null,"type":"comment","url":null},{"author":"TeMPOraL","children":[],"created_at":"2026-03-31T12:21:57.000Z","created_at_i":1774959717,"id":47586277,"options":[],"parent_id":47585954,"points":null,"story_id":47584540,"text":"Nah, it's a hallmark of your average codebase in pre-LLM era.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:47:45.000Z","created_at_i":1774957665,"id":47585954,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Ridiculous string comparisons on long chains of logic are a hallmark of vibe-coding.","title":null,"type":"comment","url":null},{"author":"smef","children":[],"created_at":"2026-03-31T12:02:04.000Z","created_at_i":1774958524,"id":47586096,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"so frustrating..","title":null,"type":"comment","url":null},{"author":"speedgoose","children":[],"created_at":"2026-03-31T12:28:24.000Z","created_at_i":1774960104,"id":47586350,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"I guess using French words is safe for now.","title":null,"type":"comment","url":null},{"author":"ozim","children":[],"created_at":"2026-03-31T12:30:19.000Z","created_at_i":1774960219,"id":47586372,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"There is no \u201estupid\u201d I often write \u201e(this is stupid|are you stupid) fix this\u201d.
And Claude was having in chain of though \u201euser is frustrated\u201d and I wrote to it I am not frustrated just testing prompt optimization where acting like one is frustrated should yield better results.","title":null,"type":"comment","url":null},{"author":"alex_duf","children":[],"created_at":"2026-03-31T12:32:40.000Z","created_at_i":1774960360,"id":47586404,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"everyone here is commenting how odd it looks to use a regexp for sentiment analysis, but it depends what they're trying to do.
It could be used as a feedback when they do A/B test and they can compare which version of the model is getting more insult than the other. It doesn't matter if the list is exhaustive or even sane, what matters is how you compare it to the other.
Perfect? no. Good and cheap indicator? maybe.","title":null,"type":"comment","url":null},{"author":"francisofascii","children":[{"author":"nananana9","children":[],"created_at":"2026-03-31T13:06:56.000Z","created_at_i":1774962416,"id":47586807,"options":[],"parent_id":47586420,"points":null,"story_id":47584540,"text":"I doubt they're all classified the same. I'd guess they're using this regex as a litmus test to check if something should be submitted at all, they can then do deeper analysis offline after the fact.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:33:42.000Z","created_at_i":1774960422,"id":47586420,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Interesting that expletives and words that are more benign like "frustrating" are all classified the same.","title":null,"type":"comment","url":null},{"author":"1970-01-01","children":[],"created_at":"2026-03-31T12:42:34.000Z","created_at_i":1774960954,"id":47586507,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Hmm.. I flag things as 'broken' often and I've been asked to rate my sessions almost daily. Now I see why.","title":null,"type":"comment","url":null},{"author":"stainablesteel","children":[],"created_at":"2026-03-31T12:54:33.000Z","created_at_i":1774961673,"id":47586665,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"i dislike LLMs going down that road, i don't want to be punished for being mean to the clanker","title":null,"type":"comment","url":null},{"author":"ccvannorman","children":[],"created_at":"2026-03-31T13:02:59.000Z","created_at_i":1774962179,"id":47586764,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"you'd better be careful wth your typos, as well","title":null,"type":"comment","url":null},{"author":"gilbetron","children":[],"created_at":"2026-03-31T13:14:51.000Z","created_at_i":1774962891,"id":47586905,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"That's undoubtedly to detect frustration signals, a useful metric/signal for UX. The UI equivalent is the user shaking their mouse around or clicking really fast.","title":null,"type":"comment","url":null},{"author":"pprotas","children":[{"author":"jeanlucas","children":[{"author":"glitch13","children":[{"author":"drstewart","children":[],"created_at":"2026-03-31T14:30:06.000Z","created_at_i":1774967406,"id":47587923,"options":[],"parent_id":47587788,"points":null,"story_id":47584540,"text":"(yes|no|maybe)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:20:11.000Z","created_at_i":1774966811,"id":47587788,"options":[],"parent_id":47587031,"points":null,"story_id":47584540,"text":"Could you share the regex you used to come up with that sentiment analysis?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:24:54.000Z","created_at_i":1774963494,"id":47587031,"options":[],"parent_id":47586937,"points":null,"story_id":47584540,"text":">Everyone is commenting how this regex is actually a master optimization move by Anthropic
No? I'd say not even 50% of the comments are positive right now.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:17:21.000Z","created_at_i":1774963041,"id":47586937,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Everyone is commenting how this regex is actually a master optimization move by Anthropic
When in reality this is just what their LLM coding agent came up with when some engineer told it to "log user frustration"","title":null,"type":"comment","url":null},{"author":"bean469","children":[],"created_at":"2026-03-31T13:28:23.000Z","created_at_i":1774963703,"id":47587071,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Curiously "clanker" is not on the list","title":null,"type":"comment","url":null},{"author":"mcv","children":[{"author":"integralid","children":[{"author":"SoftTalker","children":[{"author":"mmh0000","children":[{"author":"irishcoffee","children":[],"created_at":"2026-03-31T18:17:41.000Z","created_at_i":1774981061,"id":47591391,"options":[],"parent_id":47590418,"points":null,"story_id":47584540,"text":"What really works? Sycophancy? I think that is a bug, not a feature.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:08:28.000Z","created_at_i":1774976908,"id":47590418,"options":[],"parent_id":47588715,"points":null,"story_id":47584540,"text":"It actually works really well if you suck up to the AI.
"Please do x"
"Thank you, that works great! Please do y now."
"You're so smart!"
lol. It really works though! At least in my experience, Claude gets almost hostile or "annoyed" when I'm not nice enough to it. And I swear it purposefully acts like a "malicious genie" when I'm not nice enough. "It works, exactly like you requested, but what you requested is stupid. Let me show you how stupid you are."
But, when I'm nice, it is way more open, like "Are you sure you really want to do X? You probably want X+Y."","title":null,"type":"comment","url":null},{"author":"soiltype","children":[{"author":"alxndr","children":[],"created_at":"2026-04-02T00:29:22.000Z","created_at_i":1775089762,"id":47608532,"options":[],"parent_id":47594753,"points":null,"story_id":47584540,"text":"I think if you treat it like a coworker who you respect, it will speak to you like a coworker who respects you, but will still make some idiotic mistakes...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:18:35.000Z","created_at_i":1774999115,"id":47594753,"options":[],"parent_id":47588715,"points":null,"story_id":47584540,"text":"It seems strange to you? It's natural to how I write - intentionally avoiding politeness would be weirder to me.
But aside from that, an LLM is only a roleplayer. Treat it like an idiot that makes mistakes and it will act like one. Treat it like a coworker who you respect and it will act like one, and it will find better results.
Obviously nothing about how they act is set in stone but as a general rule this seems to me to be both wise and, in my experience, true as well.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:23:38.000Z","created_at_i":1774970618,"id":47588715,"options":[],"parent_id":47587722,"points":null,"story_id":47584540,"text":"Always seems strange to me that people say "please" and "thank you" to LLMs.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:13:57.000Z","created_at_i":1774966437,"id":47587722,"options":[],"parent_id":47587204,"points":null,"story_id":47584540,"text":"I always type "please continue". I guess being polite is not a good idea.","title":null,"type":"comment","url":null},{"author":"dostick","children":[],"created_at":"2026-03-31T14:27:24.000Z","created_at_i":1774967244,"id":47587889,"options":[],"parent_id":47587204,"points":null,"story_id":47584540,"text":"\u201cGo on\u201d works fine too","title":null,"type":"comment","url":null},{"author":"hombre_fatal","children":[],"created_at":"2026-03-31T15:10:03.000Z","created_at_i":1774969803,"id":47588503,"options":[],"parent_id":47587204,"points":null,"story_id":47584540,"text":"The only time that function is used in the code is to log it.
logEvent('tengu_input_prompt', { isNegative, isKeepGoing })Also:
// Match "continue" only if it's the entire prompt\n if (lowerInput === 'continue') {\n return true\n }\nI know I used this word two days ago when I went through three rounds of an agent telling me that it fixed three things without actually changing them.
I think starting a new session and telling it that the previous agent's work / state was terrible (so explain what happened) is pretty unremarkable. It's certainly not saying "fuck you". I think this is a little silly.","title":null,"type":"comment","url":null},{"author":"ezekg","children":[],"created_at":"2026-03-31T14:09:41.000Z","created_at_i":1774966181,"id":47587645,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Nice, "wtaf" doesn't match so I think I'm out of the dog house when the clanker hits AGI (probably).","title":null,"type":"comment","url":null},{"author":"stefanovitti","children":[],"created_at":"2026-03-31T14:19:59.000Z","created_at_i":1774966799,"id":47587786,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"so they think that everybody on earth swears only in english?","title":null,"type":"comment","url":null},{"author":"ZainRiz","children":[{"author":"indigodaddy","children":[],"created_at":"2026-03-31T14:32:15.000Z","created_at_i":1774967535,"id":47587951,"options":[],"parent_id":47587943,"points":null,"story_id":47584540,"text":"Continue?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:31:36.000Z","created_at_i":1774967496,"id":47587943,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"They also have a "keep going" keyword, literally just "continue" or "keep going", just for logging.
I've been using "resume" this whole time","title":null,"type":"comment","url":null},{"author":"joeblau","children":[{"author":"vdfs","children":[{"author":"adzm","children":[{"author":"trych","children":[],"created_at":"2026-04-01T13:42:04.000Z","created_at_i":1775050924,"id":47600749,"options":[],"parent_id":47589743,"points":null,"story_id":47584540,"text":"Smart, that way they will never know when you're shouting at them.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:21:51.000Z","created_at_i":1774974111,"id":47589743,"options":[],"parent_id":47589566,"points":null,"story_id":47584540,"text":"the string is lowercased before the regex is run, fwiw","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:12:27.000Z","created_at_i":1774973547,"id":47589566,"options":[],"parent_id":47588076,"points":null,"story_id":47584540,"text":"Did you also forget to ignore case sensitivity back then?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:42:21.000Z","created_at_i":1774968141,"id":47588076,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"We used this in 2011 at the startup I worked for. 20 positive and 20 negative words was good enough to sell Twitter "sentiment analysis" to companies like Apple, Bentley, etc...","title":null,"type":"comment","url":null},{"author":"AIorNot","children":[],"created_at":"2026-03-31T14:44:36.000Z","created_at_i":1774968276,"id":47588105,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"OMG WTF","title":null,"type":"comment","url":null},{"author":"FranOntanaya","children":[],"created_at":"2026-03-31T14:46:29.000Z","created_at_i":1774968389,"id":47588128,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"That looks a bit bare minimum, not the use of regex but rather that it's a single line with a few dozen words. You'd think they'd have a more comprehensive list somewhere and assemble or iterate the regex checks as needed.","title":null,"type":"comment","url":null},{"author":"johnfn","children":[],"created_at":"2026-03-31T14:49:40.000Z","created_at_i":1774968580,"id":47588182,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Surely "so frustrating" isn't explicit content?","title":null,"type":"comment","url":null},{"author":"nico","children":[{"author":"jollymonATX","children":[{"author":"DIVx0","children":[],"created_at":"2026-03-31T15:59:14.000Z","created_at_i":1774972754,"id":47589347,"options":[],"parent_id":47588868,"points":null,"story_id":47584540,"text":"I find when you give harsh feedback to claude it becomes "neurotic" and worthless, if "wtf" enters the chat, then you know it's time to restart or DIY.","title":null,"type":"comment","url":null},{"author":"nico","children":[],"created_at":"2026-03-31T18:46:49.000Z","created_at_i":1774982809,"id":47591733,"options":[],"parent_id":47588868,"points":null,"story_id":47584540,"text":"Not really. Most of the times it actually finally picks up on what I was telling it to do. Sometimes it takes a few tries, like 2-3 wtfs. I don\u2019t think I\u2019ve ever given it more than 3 consecutive wtfs, and that would be a lot
It\u2019s about a once a week or less event. A bit annoying sometimes, but not a deal breaker","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:32:13.000Z","created_at_i":1774971133,"id":47588868,"options":[],"parent_id":47588782,"points":null,"story_id":47584540,"text":"Did you notice a change in quality after you went foul?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:27:20.000Z","created_at_i":1774970840,"id":47588782,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"Probably a lot of my prompts have been logged then. I\u2019ve used wtf so many times I\u2019ve lost track. But I guess Claude hasn\u2019t","title":null,"type":"comment","url":null},{"author":"DIVx0","children":[],"created_at":"2026-03-31T15:57:18.000Z","created_at_i":1774972638,"id":47589310,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"oh I hope they really are paying attention. Even though I'm 100% aware that claude is a clanker, sometimes it just exhibits the most bizarre behavior that it triggers my lizard brain to react to it. That experience troubles me so much that I've mostly stopped using claude code. Claude won't even semi-reliably follow its own policies, sometimes even immediately after you confirm it knows about them.","title":null,"type":"comment","url":null},{"author":"amichal","children":[],"created_at":"2026-03-31T16:31:57.000Z","created_at_i":1774974717,"id":47589888,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"If this code is real and complete then there are no callers of those methods other than a logger line","title":null,"type":"comment","url":null},{"author":"rurp","children":[],"created_at":"2026-03-31T16:39:53.000Z","created_at_i":1774975193,"id":47590004,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"I was thinking the opposite. Using those words might be the best way to provide feedback that actually gets considered.
I've been wondering if all of these companies have some system for flagging upset responses. Those cases seem like they are far more likely than average to point to weaknesses in the model and/or potentially dangerous situations.","title":null,"type":"comment","url":null},{"author":"jacquesm","children":[],"created_at":"2026-03-31T20:15:44.000Z","created_at_i":1774988144,"id":47592862,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"George Carlin would be very pleased. They missed quite a few of the heavy seven though.","title":null,"type":"comment","url":null},{"author":"shardullavekar","children":[],"created_at":"2026-04-01T12:55:41.000Z","created_at_i":1775048141,"id":47600202,"options":[],"parent_id":47585326,"points":null,"story_id":47584540,"text":"wondering how this fares for languages other than English.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:39:01.000Z","created_at_i":1774953541,"id":47585326,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"They have an interesting regex for detecting negative sentiment in users prompt which is then logged (explicit content): https://github.com/chatgptprojects/claude-code/blob/642c7f94...
I guess these words are to be avoided...","title":null,"type":"comment","url":null},{"author":"mesmertech","children":[{"author":"mesmertech","children":[{"author":"mattmanser","children":[{"author":"mr_00ff00","children":[],"created_at":"2026-03-31T13:43:40.000Z","created_at_i":1774964620,"id":47587300,"options":[],"parent_id":47586135,"points":null,"story_id":47584540,"text":"It\u2019s trained on mostly internet content, right?
If it learned language based on how the internet talks, then the best way to communicate is using similar language.","title":null,"type":"comment","url":null},{"author":"fermentation","children":[],"created_at":"2026-03-31T15:40:21.000Z","created_at_i":1774971621,"id":47589019,"options":[],"parent_id":47586135,"points":null,"story_id":47584540,"text":"This is claude writing code for itself. It talks like this to itself when you ask it to make prompts.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:06:15.000Z","created_at_i":1774958775,"id":47586135,"options":[],"parent_id":47585393,"points":null,"story_id":47584540,"text":"One tengentially interesting thing about that is how THEY talk to Claude.
"Don't blow your cover"
Interesting to see them be so informal and use an idiom to a computer.
And using capitals for emphasis.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:46:07.000Z","created_at_i":1774953967,"id":47585393,"options":[],"parent_id":47585377,"points":null,"story_id":47584540,"text":"Also saw this on twitter earlier, thought someone was just making a fake hype post thing. But turns out to be an actual prompt for capybara huh: https://github.com/chatgptprojects/claude-code/blob/642c7f94...","title":null,"type":"comment","url":null},{"author":"mesmertech","children":[{"author":"nunez","children":[],"created_at":"2026-03-31T15:06:57.000Z","created_at_i":1774969617,"id":47588446,"options":[],"parent_id":47585719,"points":null,"story_id":47584540,"text":"They even leaked their April Fool\u2019s fun. Brutal!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:18:54.000Z","created_at_i":1774955934,"id":47585719,"options":[],"parent_id":47585377,"points":null,"story_id":47584540,"text":"turns out its for an April fools tomorrow: https://x.com/mesmerlord/status/2038938888178135223","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:43:50.000Z","created_at_i":1774953830,"id":47585377,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Was searching for the rumored Mythos/Capybara release, and what even is this file? https://github.com/chatgptprojects/claude-code/blob/642c7f94...","title":null,"type":"comment","url":null},{"author":"Diablo556","children":[{"author":"derwiki","children":[{"author":"infinitezest","children":[],"created_at":"2026-03-31T13:17:37.000Z","created_at_i":1774963057,"id":47586941,"options":[],"parent_id":47586528,"points":null,"story_id":47584540,"text":"And they can't hear you from under the enormous pile of debt they're fighting to overcome. Maybe try again in 2028.","title":null,"type":"comment","url":null},{"author":"flexagoon","children":[],"created_at":"2026-03-31T18:25:31.000Z","created_at_i":1774981531,"id":47591481,"options":[],"parent_id":47586528,"points":null,"story_id":47584540,"text":"> redefining what SWE means
Redefining the "SW" to stand for "slopware"?","title":null,"type":"comment","url":null},{"author":"lqstuart","children":[],"created_at":"2026-03-31T18:54:42.000Z","created_at_i":1774983282,"id":47591827,"options":[],"parent_id":47586528,"points":null,"story_id":47584540,"text":"you mean the $5 billion they've generated off of the $73 billion they've raised?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:43:43.000Z","created_at_i":1774961023,"id":47586528,"options":[],"parent_id":47585456,"points":null,"story_id":47584540,"text":"I don\u2019t think they can hear you over the billions of dollars they are generating, and definitely not over them redefining what SWE means.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:52:43.000Z","created_at_i":1774954363,"id":47585456,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"haha.. Anthropic need to hire fixer from vibecodefixers.com to fix all that messy code..lol","title":null,"type":"comment","url":null},{"author":"jedisct1","children":[{"author":"prmoustache","children":[],"created_at":"2026-03-31T11:03:53.000Z","created_at_i":1774955033,"id":47585573,"options":[],"parent_id":47585468,"points":null,"story_id":47584540,"text":"It is a client running on an interpreted language your own computer, there is nothing to secure or hide as source is provided to you already.","title":null,"type":"comment","url":null},{"author":"prmoustache","children":[{"author":"jedisct1","children":[],"created_at":"2026-03-31T11:12:16.000Z","created_at_i":1774955536,"id":47585643,"options":[],"parent_id":47585577,"points":null,"story_id":47584540,"text":"It was heavily obfuscated, keeping users in the dark about what they\u2019re installing and running.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:04:49.000Z","created_at_i":1774955089,"id":47585577,"options":[],"parent_id":47585468,"points":null,"story_id":47584540,"text":"It is a client running on an interpreted language your own computer, there is nothing to secure or hide as source was provided to you already or am I mistaking?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:53:48.000Z","created_at_i":1774954428,"id":47585468,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"It shows that a company you and your organization are trusting with your data, and allowing full control over your devices 24/7, is failing to properly secure its own software.
It's a wake up call.","title":null,"type":"comment","url":null},{"author":"mohsen1","children":[{"author":"phtrivier","children":[{"author":"konart","children":[],"created_at":"2026-03-31T11:53:51.000Z","created_at_i":1774958031,"id":47586027,"options":[],"parent_id":47585834,"points":null,"story_id":47584540,"text":"Can't we have generated / llm generated code to be more human maintainable?","title":null,"type":"comment","url":null},{"author":"grey-area","children":[{"author":"olmo23","children":[{"author":"AlexCoventry","children":[],"created_at":"2026-04-02T01:38:45.000Z","created_at_i":1775093925,"id":47609024,"options":[],"parent_id":47587186,"points":null,"story_id":47584540,"text":"> due to the way round-tripping tool-calls work, splitting code up into multiple files is counter-productive.
Can you expand on that?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:36:24.000Z","created_at_i":1774964184,"id":47587186,"options":[],"parent_id":47586061,"points":null,"story_id":47584540,"text":"I agree the functions in a file should probably be reasonably-sized.
It's also interesting to note that due to the way round-tripping tool-calls work, splitting code up into multiple files is counter-productive. You're better off with a single large file.","title":null,"type":"comment","url":null},{"author":"zer00eyz","children":[{"author":"grey-area","children":[],"created_at":"2026-04-02T05:09:27.000Z","created_at_i":1775106567,"id":47610201,"options":[],"parent_id":47592464,"points":null,"story_id":47584540,"text":"I think humans are pretty good at it with small teams and the right structure. There are definitely dysfunctional orgs as you describe where humans produce garbage code yes. I blame the org for that, not the humans.
As to what defines clarity, yes of course, like the word quality this is very hard to define, but we can certainly recognise when it was not considered.
I think it is a goal worth striving for though, and abandoning code standards because we now have AI helpers is stupid and self-defeating, even if we think they are very capable and will improve.
The end of history has not in fact arrived with generative AI, we still have to maintain software after.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:46:55.000Z","created_at_i":1774986415,"id":47592464,"options":[],"parent_id":47586061,"points":null,"story_id":47584540,"text":"> independently work on a large codebase
Im not sure that Humans are great at this either. Think about how we use frameworks and have complex supply chains... we sort of get "good enough" at what we need to do and pray a lot that everything else keeps working and that our tooling (things like artifactory) save us from supply chain attacks. Or we just run piles of old, outdated code because "it works". I cant tell you how many micro services I have seen that are "just fine" but no one in the current org has ever read a line of what's in them, and the people who wrote them left ages ago.
> clarity too
Yes, but define clarity!
I recently had the pleasure of fixing a chunk of code that was part of a data pipeline. It was an If/elseif/elseif structure... where the final two states were fairly benign and would have been applicable in 99 percent of cases. Everything else was to deal with the edge cases!
I had an idea of where the issue was, but I didn't understand how the code ended up in the state it was in... Blame -> find the commit message (references ticket) -> find the Jira ticket (references sales force) -> find the original customer issue in salesforce, read through the whole exchange there.
A two line comment could have spared me all that work, to get to what amounted to a dead simple fix. The code was absolutely clear, but without the "why" portion of the context I likely would have created some sort of regression, that would have passed the good enough testing that was there.
I re-wrote a portion of the code (expanding variable names) - that code is now less "scannable" and more "readable" (different types of clarity). Dropped in comments: a few sentences of explaining, and references to the tickets. Went and updated tests, with similar notes.
Meanwhile, elsewhere (other code base, other company), that same chain is broken... the "bug tracking system" that is referenced in the commit messages there no longer exists.
I have a friend who, every time he updates his dev env, he calls me to report that he "had to go update the wiki again!" Because someone made a change and told every one in a slack message. Here is yet another vast repository of degrading, unsearchable and unusable tribal knowledge embedded in so many organizations out there.
Don't even get me started on the project descriptions/goals/tasks that amount to pantomime a post-it notes, absent of any sort of genuine description.
Lack of clarity is very much also a lack of "context" in situ problem.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:57:01.000Z","created_at_i":1774958221,"id":47586061,"options":[],"parent_id":47585834,"points":null,"story_id":47584540,"text":"LLMs are so so far away from being able to independently work on a large codebase, and why would they not benefit from modularity and clarity too?","title":null,"type":"comment","url":null},{"author":"Bayko","children":[{"author":"yoz-y","children":[],"created_at":"2026-03-31T13:58:57.000Z","created_at_i":1774965537,"id":47587516,"options":[],"parent_id":47586152,"points":null,"story_id":47584540,"text":"The jury on this one is still out.","title":null,"type":"comment","url":null},{"author":"phtrivier","children":[{"author":"heavyset_go","children":[],"created_at":"2026-03-31T16:26:43.000Z","created_at_i":1774974403,"id":47589809,"options":[],"parent_id":47588443,"points":null,"story_id":47584540,"text":"If that's the case then that's dumb","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:06:46.000Z","created_at_i":1774969606,"id":47588443,"options":[],"parent_id":47586152,"points":null,"story_id":47584540,"text":"Sorry, I was not clear enough.
My first word was litteraly "Yes", so I agree that a function like this is a maintenance nightmare for a human. \nAnd, sure, the code might not be "optimized" for the LLM, or token efficiency.
However, to try and make my point clearer: it's been reported that anthropic has "some developpers won't don't write code" [1].
I have no inside knowledge, but it's possible, by extension, to assume that some parts of their own codebase are "maintained" mostly by LLMs themselves.
If you push this extension, then, the code that is generated only has to be "readable" to:
* the next LLM that'll have to touch it
* the compiler / interpreter that is going to compile / run it.
In a sense (and I know this is a stretch, and I don't want to overdo the analogy), are we, here, judging a program quality by reading something more akin to "the x86 asm outputed by the compiler", rather than the "source code" - which in this case, is "english prompts", hidden somewhere in the claude code session of a developper ?
Just speculating, obviously. My org is still very much more cautious, and mandating people to have the same standard for code generated by LLM as for code generated by human ; and I agree with that.
I would _not_ want to debug the function described by the commentor.
So I'm still very much on the "claude as a very fast text editor" side, but is it unreasonnable to assume that anthropic might be further on the "claude as a compiler for english" side ?
[1] https://www.reddit.com/r/ArtificialInteligence/comments/1s7j...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:08:48.000Z","created_at_i":1774958928,"id":47586152,"options":[],"parent_id":47585834,"points":null,"story_id":47584540,"text":"Ye I honestly don't understand his comment. Is it bad code writing? Pre 2026? Sure. In 2026. Nope. Is it going to be a headache for some poor person on oncall? Yes. But then again are you "supposed" to go through every single line in 2026? Again no. I hate it. But the world is changing and till the bubble pops this is the new norm","title":null,"type":"comment","url":null},{"author":"mdavid626","children":[{"author":"jen20","children":[{"author":"mvfc","children":[],"created_at":"2026-04-08T11:37:08.000Z","created_at_i":1775648228,"id":47688783,"options":[],"parent_id":47590840,"points":null,"story_id":47584540,"text":"The company does pay per usage for something though, be it data center usage, be it water and power for their own data centers.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:35:09.000Z","created_at_i":1774978509,"id":47590840,"options":[],"parent_id":47586207,"points":null,"story_id":47584540,"text":"I very much doubt Anthropic devs are metered, somehow.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:14:55.000Z","created_at_i":1774959295,"id":47586207,"options":[],"parent_id":47585834,"points":null,"story_id":47584540,"text":"Oh boy, you couldn't be more wrong. If something, LLM-s need MORE readable code, not less. Do you want to burn all your money in tokens?","title":null,"type":"comment","url":null},{"author":"mrbungie","children":[{"author":"phtrivier","children":[],"created_at":"2026-03-31T15:11:38.000Z","created_at_i":1774969898,"id":47588523,"options":[],"parent_id":47586219,"points":null,"story_id":47584540,"text":"Oh, do we agree on that. I never said it was "smart" - I just had a theory that would explain why such code could exist (see my longer answer below).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:16:11.000Z","created_at_i":1774959371,"id":47586219,"options":[],"parent_id":47585834,"points":null,"story_id":47584540,"text":"Can't wait to have LLM generated physical objects that explode on you face and no engineer can fix.","title":null,"type":"comment","url":null},{"author":"ulrikrasmussen","children":[{"author":"dntrkv","children":[],"created_at":"2026-04-01T08:26:02.000Z","created_at_i":1775031962,"id":47598304,"options":[],"parent_id":47592504,"points":null,"story_id":47584540,"text":"Massive tests files are almost as bad a massive function.
Scrolling through a 3k line test suite with multiple levels of nesting trying to figure out which cases are covered is a fucking pain in the ass.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:49:48.000Z","created_at_i":1774986588,"id":47592504,"options":[],"parent_id":47585834,"points":null,"story_id":47584540,"text":"Unit testing is much much harder when you have functions spanning thousands of lines and no abstractions. You have to white box test everything to ensure that you hit all code paths, and it is much more expensive to maintain such tests, both as a human and LLM. I don't think this can be ignored just because LLMs are writing the code.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:33:43.000Z","created_at_i":1774956823,"id":47585834,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"Yes, if it was made for human comprehension or maintenance.
If it's entirely generated / consumed / edited by an LLM, arguably the most important metric is... test coverage, and that's it ?","title":null,"type":"comment","url":null},{"author":"mohsen1","children":[{"author":"almostdeadguy","children":[{"author":"mrcwinn","children":[{"author":"almostdeadguy","children":[],"created_at":"2026-03-31T18:22:32.000Z","created_at_i":1774981352,"id":47591454,"options":[],"parent_id":47591373,"points":null,"story_id":47584540,"text":"Yes, anthropic is not the only company in the world with some shitty code, and yet I feel no pangs of guilt over laughing about it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:16:33.000Z","created_at_i":1774980993,"id":47591373,"options":[],"parent_id":47589616,"points":null,"story_id":47584540,"text":"You fail to mention the prior decades of really bad software engineers writing awful code -- off of which these models trained.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:14:42.000Z","created_at_i":1774973682,"id":47589616,"options":[],"parent_id":47586261,"points":null,"story_id":47584540,"text":"A defining work of the "just vibes" era.","title":null,"type":"comment","url":null},{"author":"visarga","children":[],"created_at":"2026-03-31T16:52:22.000Z","created_at_i":1774975942,"id":47590189,"options":[],"parent_id":47586261,"points":null,"story_id":47584540,"text":"Claude Code says thank you for reporting, I bet they will scan this chat to see what bugs they need to fix asap.","title":null,"type":"comment","url":null},{"author":"NooneAtAll3","children":[{"author":"clankerbad","children":[{"author":"komali2","children":[],"created_at":"2026-04-01T14:08:09.000Z","created_at_i":1775052489,"id":47601110,"options":[],"parent_id":47596926,"points":null,"story_id":47584540,"text":"wl-copy is a program to put text into the system clipboard if you're on a wayland-based system (so you can ctrl-v paste it somewhere else). Imagine like, cat ~/.ssh/whatever | wl-copy and then pasting into github or something.
xclip is the same for X based systems.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T04:45:08.000Z","created_at_i":1775018708,"id":47596926,"options":[],"parent_id":47596077,"points":null,"story_id":47584540,"text":"Looks like it tries wl-copy, then tries xclip and then tries xsel. I have no idea what those are but google says it's for Wayland, so, I think it's a linux function trying to copy to clipboard? I think their problem is with the use of '.then(...=>...)' since there doesn't seem to be a way to tell each function that the nested ones actually finished.","title":null,"type":"comment","url":null},{"author":"dolmen","children":[],"created_at":"2026-04-01T09:51:49.000Z","created_at_i":1775037109,"id":47598819,"options":[],"parent_id":47596077,"points":null,"story_id":47584540,"text":"It looks like a search for the command line tool available to send content to the clipboard.
Can't tell if that obfuscated code works though.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T02:31:43.000Z","created_at_i":1775010703,"id":47596077,"options":[],"parent_id":47586261,"points":null,"story_id":47584540,"text":"what does that even do?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:20:43.000Z","created_at_i":1774959643,"id":47586261,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"here's another gem. src/ink/termio/osc.ts:192\u2013210
void execFileNoThrow('wl-copy', [], opts).then(r => {\n if (r.code === 0) { linuxCopy = 'wl-copy'; return }\n void execFileNoThrow('xclip', ...).then(r2 => {\n if (r2.code === 0) { linuxCopy = 'xclip'; return }\n void execFileNoThrow('xsel', ...).then(r3 => {\n linuxCopy = r3.code === 0 ? 'xsel' : null\n })\n })\n })\n\nClose.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:35:05.000Z","created_at_i":1774964105,"id":47587164,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"I'm sure this is no surprise to anyone who has used CC for a while. This is the source of so many bugs. I would say "open bugs" but Anthropic auto-closes bugs that don't have movement on them in like 60 days.","title":null,"type":"comment","url":null},{"author":"mohsen1","children":[],"created_at":"2026-03-31T13:38:13.000Z","created_at_i":1774964293,"id":47587210,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"it's the `runHeadlessStreaming` function btw","title":null,"type":"comment","url":null},{"author":"acedTrex","children":[],"created_at":"2026-03-31T13:55:10.000Z","created_at_i":1774965310,"id":47587445,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"Well, literally no one has ever accused anthropic of having even half way competent engineers. They are akin to monkeys whacking stuff with a stick.","title":null,"type":"comment","url":null},{"author":"ykonstant","children":[{"author":"dwa3592","children":[{"author":"dwa3592","children":[{"author":"jamiek88","children":[],"created_at":"2026-03-31T22:50:49.000Z","created_at_i":1774997449,"id":47594513,"options":[],"parent_id":47594309,"points":null,"story_id":47584540,"text":"Because it adds nothing to the conversation and has a Reddit vibe and that goes down like a lead balloon in these here parts, cowboy.","title":null,"type":"comment","url":null},{"author":"gsinclair","children":[],"created_at":"2026-04-01T03:30:40.000Z","created_at_i":1775014240,"id":47596475,"options":[],"parent_id":47594309,"points":null,"story_id":47584540,"text":"Take a look at the site guidelines.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:28:09.000Z","created_at_i":1774996089,"id":47594309,"options":[],"parent_id":47587837,"points":null,"story_id":47584540,"text":"i wonder why 'lmao' gets downvoted.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:23:32.000Z","created_at_i":1774967012,"id":47587837,"options":[],"parent_id":47587762,"points":null,"story_id":47584540,"text":"lmao","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:18:00.000Z","created_at_i":1774966680,"id":47587762,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":""That's Larry; he does most of the work around here."","title":null,"type":"comment","url":null},{"author":"DustinBrett","children":[],"created_at":"2026-03-31T14:41:54.000Z","created_at_i":1774968114,"id":47588068,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":""You can get Claude to split that up"","title":null,"type":"comment","url":null},{"author":"keeganpoppen","children":[{"author":"lanbin","children":[],"created_at":"2026-03-31T16:13:49.000Z","created_at_i":1774973629,"id":47589594,"options":[],"parent_id":47588866,"points":null,"story_id":47584540,"text":"I see. They got unlimited tokens, right?","title":null,"type":"comment","url":null},{"author":"knome","children":[{"author":"eclipxe","children":[{"author":"fl4regun","children":[{"author":"lijok","children":[{"author":"hallway_monitor","children":[{"author":"GrinningFool","children":[{"author":"dntrkv","children":[],"created_at":"2026-04-01T08:19:20.000Z","created_at_i":1775031560,"id":47598263,"options":[],"parent_id":47593345,"points":null,"story_id":47584540,"text":"I have been seeing this problem building over the last year. LLM generated logic being tested by massive LLM generated tests.
Everyone just goes overboard with the tests since you can easily just tell the LLM to expand on the suite. So you end up with a massive test suite that looks very thorough and is less likely to be scrutinized.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:53:52.000Z","created_at_i":1774990432,"id":47593345,"options":[],"parent_id":47592451,"points":null,"story_id":47584540,"text":"But unit and integration tests generally only catch the things you can think of. That leaves a lot of unexplored space in which things can go wrong.
Separately, but related - if you offload writing of the tests and writing of the code, how does anybody know what they have other than green tests and coverage numbers?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:46:08.000Z","created_at_i":1774986368,"id":47592451,"options":[],"parent_id":47591959,"points":null,"story_id":47584540,"text":"Exactly.... -> Unit tests. Integration tests. UI tests. This is how code should be verified no matter the author. Just today I told my team we should not be reading every line of LLM code. Understand the pattern. Read the interesting / complex parts. Read the tests.","title":null,"type":"comment","url":null},{"author":"fl4regun","children":[],"created_at":"2026-04-02T18:41:03.000Z","created_at_i":1775155263,"id":47618455,"options":[],"parent_id":47591959,"points":null,"story_id":47584540,"text":"if you are asking me how you *guarantee* there is not a single possible exploit in your code, you can't do that. But you can do your best and learn about common pitfalls and be reasonably competent. Just because you can't do the former doesn't mean the latter is useless.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:04:06.000Z","created_at_i":1774983846,"id":47591959,"options":[],"parent_id":47591040,"points":null,"story_id":47584540,"text":"How do you with classic code?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:49:24.000Z","created_at_i":1774979364,"id":47591040,"options":[],"parent_id":47590449,"points":null,"story_id":47584540,"text":"Is this a serious question? If you are handling sensitive information how do you confirm your application is secure and won't leak or expose information to people who shouldn't know it?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:10:28.000Z","created_at_i":1774977028,"id":47590449,"options":[],"parent_id":47589657,"points":null,"story_id":47584540,"text":"Why?","title":null,"type":"comment","url":null},{"author":"jen20","children":[{"author":"batshit_beaver","children":[{"author":"bdangubic","children":[{"author":"batshit_beaver","children":[{"author":"bdangubic","children":[],"created_at":"2026-04-02T11:18:37.000Z","created_at_i":1775128717,"id":47612866,"options":[],"parent_id":47592957,"points":null,"story_id":47584540,"text":"I find this relationship fascinating. since the OSS vast majority of the developers will not hesitate to pull in library X or framework Y knowing really nothing about it, who are developers, what is the quality of it, what is their release process, qa etc etc... The first thing I do now as a "senior" for decades when I get approached with "we should consider using ____" is to send them to their issues page ( e.g. https://github.com/oven-sh/bun/issues ) and then be like "spend 60-90 minutes minimum here reviewing the issues - then come back and tell me whether or not the inclusion of this is something we should consider." and yet, now with LLMs there are sooooooooo many comments on HN like "oh they must be supervised, who knows what they will be doing etc..." - gotta supervise them but some mate in Boise is all good, hopefully someone else will review his stuff that is going into your next release ...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:22:59.000Z","created_at_i":1774988579,"id":47592957,"options":[],"parent_id":47591701,"points":null,"story_id":47584540,"text":"I'd trust that dude over professional leetcoders any day.
But you're right that trust is a complicated thing and often misplaced. I think as an industry we're always reevaluating our relationship with OSS, and I'm sure LLMs will affect this relationship in some way. It's too early to tell.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:44:20.000Z","created_at_i":1774982660,"id":47591701,"options":[],"parent_id":47591233,"points":null,"story_id":47584540,"text":"except they are vibe-or-not coded by some dude in Reno NV who wouldn\u2019t pass a phone screen where you work","title":null,"type":"comment","url":null},{"author":"dolmen","children":[],"created_at":"2026-04-01T16:36:43.000Z","created_at_i":1775061403,"id":47603181,"options":[],"parent_id":47591233,"points":null,"story_id":47584540,"text":"Stop trusting any dependency now.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:04:34.000Z","created_at_i":1774980274,"id":47591233,"options":[],"parent_id":47590826,"points":null,"story_id":47584540,"text":"Because you trust that your dependencies are not vibe coded and have been reviewed by humans.","title":null,"type":"comment","url":null},{"author":"almostdeadguy","children":[],"created_at":"2026-03-31T18:20:04.000Z","created_at_i":1774981204,"id":47591427,"options":[],"parent_id":47590826,"points":null,"story_id":47584540,"text":"There's a reputational filtering that happens when using dependencies. Stars, downloads, last release, who the developer is, etc.
Yeah we get supply chain attacks (like the axios thing today) with dependencies, but on the whole I think this is much safer than YOLO git-push-force-origin-main-ing some vibe-coded trash that nobody has ever run before.
I also think this isn't really true for the FAANGs, who ostensibly vendor and heavily review many of their dependencies because of the potential impacts they face from them being wrong. For us small potatoes I think "reviewing the code in your repository" is a common sense quality check.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:34:21.000Z","created_at_i":1774978461,"id":47590826,"options":[],"parent_id":47589657,"points":null,"story_id":47584540,"text":"(I mostly agree with you, but) devils advocate: most people already do that with dependencies, so why not move the line even further up?","title":null,"type":"comment","url":null},{"author":"ramraj07","children":[{"author":"Uhhrrr","children":[{"author":"jmalicki","children":[{"author":"NothingAboutAny","children":[],"created_at":"2026-04-01T05:28:04.000Z","created_at_i":1775021284,"id":47597153,"options":[],"parent_id":47595799,"points":null,"story_id":47584540,"text":"yes but if a dev pushes a line of code that wipes the accounts of millions of users at a fintech, the dev will get fired but the CEO will get sued into oblivion.\nif the agent isn't responsible, you HAVE to be, cause angry people wont listen to "it's no ones fault your money is gone"","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T01:50:17.000Z","created_at_i":1775008217,"id":47595799,"options":[],"parent_id":47594960,"points":null,"story_id":47584540,"text":"Is the CEO responsible for a company's financial performance? Do they review every line of code the company writes?
It is more irresponsible to spend the time reviewing all of the code rather than spending that time on things with bigger levers for satisfying your customers.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:45:15.000Z","created_at_i":1775000715,"id":47594960,"options":[],"parent_id":47594523,"points":null,"story_id":47584540,"text":"If you don't review what the product does, you are irresponsible for the product.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:51:55.000Z","created_at_i":1774997515,"id":47594523,"options":[],"parent_id":47589657,"points":null,"story_id":47584540,"text":"You are still responsible for the product; the code has stopped being what defines the product.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:16:54.000Z","created_at_i":1774973814,"id":47589657,"options":[],"parent_id":47588866,"points":null,"story_id":47584540,"text":">there is no point in code-reviewing ai-generated code
the idea that you should just blindly trust code you are responsible for without bothering to review it is ludicrous.","title":null,"type":"comment","url":null},{"author":"Salgat","children":[],"created_at":"2026-03-31T16:26:47.000Z","created_at_i":1774974407,"id":47589812,"options":[],"parent_id":47588866,"points":null,"story_id":47584540,"text":"While the technology is young, bugs are to be expected, but I'm curious what happens when their competitors' mature their product, clean up the bugs and stabilize it, while Claude is still kept in this trap where a certain number of bugs and issues are just a constant fixture due to vibe coding. But hey, maybe they really do achieve AGI and get over the limitations of vibe coding without human involvement.","title":null,"type":"comment","url":null},{"author":"lqstuart","children":[],"created_at":"2026-03-31T18:41:51.000Z","created_at_i":1774982511,"id":47591668,"options":[],"parent_id":47588866,"points":null,"story_id":47584540,"text":"yes, because who ever heard of an AI leaking passwords or API keys into source code","title":null,"type":"comment","url":null},{"author":"slashdave","children":[],"created_at":"2026-04-01T05:39:15.000Z","created_at_i":1775021955,"id":47597215,"options":[],"parent_id":47588866,"points":null,"story_id":47584540,"text":"> it may be sucky aesthetically
It's not a matter of being pretty, but of being robust and maintainable.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:32:05.000Z","created_at_i":1774971125,"id":47588866,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"the claude code team ethos, as far as i\u2019ve been lead to understand\u2014 which i agree with, mind you\u2014 is that there is no point in code-reviewing ai-generated code\u2026 simply update your spec(s) and regenerate. it is just a completely different way of interacting with the world. but it clearly works for them, so people throwing up their hands should at least take notice of the fact that they are absolutely not competing with traditional code along traditional lines. it may be sucky aesthetically, but they have proven from their velocity that it can be extremely effective. welcome to the New World Order, my friend.","title":null,"type":"comment","url":null},{"author":"jollymonATX","children":[],"created_at":"2026-03-31T15:53:56.000Z","created_at_i":1774972436,"id":47589258,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"Maybe going slow is a feature for them? A kind of rate limit by bad code way to controlling overall throughput.","title":null,"type":"comment","url":null},{"author":"0xbadcafebee","children":[{"author":"FuckButtons","children":[],"created_at":"2026-03-31T16:47:21.000Z","created_at_i":1774975641,"id":47590117,"options":[],"parent_id":47589632,"points":null,"story_id":47584540,"text":"I\u2019ve heard this take before, but if you\u2019ve spent any time with llm\u2019s I don\u2019t understand how your take can be: \u201cI should just let this thing that makes mistakes all the time and seems oblivious to the complexity it\u2019s creating because it only observes small snippets out of context make it\u2019s own decisions about architecture, this is just how it does things and I shouldn\u2019t question it.\u201d","title":null,"type":"comment","url":null},{"author":"meffmadd","children":[],"created_at":"2026-03-31T16:48:35.000Z","created_at_i":1774975715,"id":47590130,"options":[],"parent_id":47589632,"points":null,"story_id":47584540,"text":"I think this view assumes no human will/should ever read the code. This is considered bad practice because someone else will not understand the code as well whether written by a human or agent. Unless 0% human oversight is needed anymore agents should still code like us.","title":null,"type":"comment","url":null},{"author":"zarzavat","children":[],"created_at":"2026-03-31T17:20:16.000Z","created_at_i":1774977616,"id":47590598,"options":[],"parent_id":47589632,"points":null,"story_id":47584540,"text":"I don't buy this. Claude doesn't usually have any issues understanding my code. It has tons of issues understanding its code.
The difference between my code and Claude's code is that when my code is getting too complex to fit in my head, I stop and refactor it, since for me understanding the code is a prerequisite for writing code.
Claude, on the other hand, will simply keep generating code well past the point when it has lost comprehension. I have to stop, revert, and tell it to do it again with a new prompt.
If anything, Claude has a greater need for structure than me since the entire task has to fit in the relatively small context window.","title":null,"type":"comment","url":null},{"author":"crakhamster01","children":[{"author":"0xbadcafebee","children":[{"author":"svachalek","children":[{"author":"0xbadcafebee","children":[],"created_at":"2026-04-01T02:41:06.000Z","created_at_i":1775011266,"id":47596140,"options":[],"parent_id":47593635,"points":null,"story_id":47584540,"text":"I don't think there's intention. And yes, its output is defined by its limits. But it's not just the context, is it? Their coding style is, above all else, a result of an algorithm and input. The training data, the reinforcement, the model design, the tuning, the prompt, the context. Change any one of those things and the code changes. They are a system, like an ecosystem. Let water flow and it finds its own path. But try to dam it and it creates unintended consequences. I think what we're going to find is some of our rules apply more to a human world than an LLM world.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:22:49.000Z","created_at_i":1774992169,"id":47593635,"options":[],"parent_id":47592848,"points":null,"story_id":47584540,"text":"They write code differently but that doesn't mean that's the kind of code they prefer to read. Don't ascribe too much intention to a stochastic process.
Their coding style is above all else a symptom of their very limited context window and complete amnesia for anything that's not in the window.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:14:54.000Z","created_at_i":1774988094,"id":47592848,"options":[],"parent_id":47591312,"points":null,"story_id":47584540,"text":"> They became popular because teams across the industry analyzed code responsible for bugs/SEVs, and all found high correlation between these metrics and shipping defects.
Yes, based on research of human code. LLMs write code differently. We should question whether the human research applies to LLMs at all. (You wouldn't take your assumptions about chimp research and apply them to parrots without confirming first)
> I think the worst outcome would be throwing out our collective wisdom because the AI labs tell us to.
We don't have to throw it out. But our current use of LLMs are a dramatic change from what came before. We should be questioning our assumptions and traditions that come from a different way of working and intelligence. Humans have a habit of trying to force things to be how they think they should be, rather than allowing them to grow organically, when the latter is often better for a system we don't yet understand.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:11:28.000Z","created_at_i":1774980688,"id":47591312,"options":[],"parent_id":47589632,"points":null,"story_id":47584540,"text":"> One thing we know for sure is LLMs write code differently than we do.
Kind of. One thing we do know for certain is that LLMs degrade in performance with context length. You will undoubtedly get worse results if the LLM has to reason through long functions and high LOC files. You might get to a working state eventually, but only after burning many more tokens than if given the right amount of context.
> The worst outcome I can imagine would be forcing them to code exactly like we do.
You're treating "code smells" like cyclomatic complexity as something that is stylistic preference, but these best practices are backed by research. They became popular because teams across the industry analyzed code responsible for bugs/SEVs, and all found high correlation between these metrics and shipping defects.
Yes, coding standards should evolve, but... that's not saying anything new. We've been iterating on them for decades now.
I think the worst outcome would be throwing out our collective wisdom because the AI labs tell us to. It might be good to question who stands to benefit when LLMs aren't leveraged efficiently.","title":null,"type":"comment","url":null},{"author":"tomcam","children":[{"author":"dpark2026","children":[],"created_at":"2026-04-01T02:27:58.000Z","created_at_i":1775010478,"id":47596053,"options":[],"parent_id":47594881,"points":null,"story_id":47584540,"text":"I'm with you there man..","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:34:43.000Z","created_at_i":1775000083,"id":47594881,"options":[],"parent_id":47589632,"points":null,"story_id":47584540,"text":"This answer blew my mind. It's making me think in a very different way.","title":null,"type":"comment","url":null},{"author":"xpe","children":[],"created_at":"2026-04-01T01:39:33.000Z","created_at_i":1775007573,"id":47595734,"options":[],"parent_id":47589632,"points":null,"story_id":47584540,"text":"Weird and inscrutable can be good: think genetic algorithms [1] such as antenna optimization for EM radiation [2]. But I like my source code on the intelligible side.
[1] https://www.nature.com/articles/s41598-023-35470-4/figures/2
[2] https://jamessealesmith.github.io/img/antenna/ant_struct.png","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:15:45.000Z","created_at_i":1774973745,"id":47589632,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"> This should be at minimum 8\u201310 separate modules.
Can't really say that for sure. The way humans structure code isn't some ideal best possible state of computer code, it's the ideal organization of computer code for human coders.
Nesting and cyclomatic complexity are indicators ("code smells"). They aren't guaranteed to lead to worse outcomes. If you have a function with 12 levels of nesting, but in each nest the first line is 'return true', you actually have 1 branch. If 2 of your 486 branch points are hit 99.999% of the time, the code is pretty dang efficient. You can't tell for sure if a design is actually good or bad until you run it a lot.
One thing we know for sure is LLMs write code differently than we do. They'll catch incredibly hard bugs while making beginner mistakes. I think we need a whole new way of analyzing their code. Our human programming rules are qualitative because it's too hard to prove if an average program does what we want. I think we need a new way to judge LLM code.
The worst outcome I can imagine would be forcing them to code exactly like we do. It just reinforces our own biases, and puts in the same bugs that we do. Vibe coding is a new paradigm, done by a new kind of intelligence. As we learn how to use it effectively, we should let the process of what works develop naturally. Evolution rather than intelligent design.","title":null,"type":"comment","url":null},{"author":"epolanski","children":[],"created_at":"2026-03-31T17:58:39.000Z","created_at_i":1774979919,"id":47591159,"options":[],"parent_id":47585511,"points":null,"story_id":47584540,"text":"Hmmm it's likely they have found that it works better for LLMs that need to operate on it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T10:57:31.000Z","created_at_i":1774954651,"id":47585511,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"src/cli/print.ts
This is the single worst function in the codebase by every metric:
![https://jamessealesmith.github.io/img/antenna/ant_struct.png](\"https://jamessealesmith.github.io/img/antenna/ant_struct.png\"){kind=link}
- 3,167 lines long (the file itself is 5,594 lines)\n - 12 levels of nesting at its deepest\n - ~486 branch points of cyclomatic complexity\n - 12 parameters + an options object with 16 sub-properties\n - Defines 21 inner functions and closures\n - Handles: agent run loop, SIGINT, rate-limits, AWS auth, MCP lifecycle, plugin install/refresh, worktree bridging, team-lead polling (while(true) inside), control message dispatch (dozens of types), model switching, turn interruption\n recovery, and more\nThere is _a lot_ of moat. Claude subscriptions are limited to Claude Code. There are proxies to impersonate Claude Code specifically for this, but Anthropic has a number of fingerprinting measures both client and server side to flag and ban these.
With the release of this source code, Anthropic basically lost the lock-in game, any proxy can now perfectly mimic Claude Code.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:06:02.000Z","created_at_i":1774955162,"id":47585590,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Gemini CLI and Codex are open source anyway. I doubt there was much of a moat there anyway. The cool kids are using things like https://pi.dev/ anyway.","title":null,"type":"comment","url":null},{"author":"BoppreH","children":[{"author":"vips7L","children":[{"author":"silversmith","children":[],"created_at":"2026-03-31T18:26:51.000Z","created_at_i":1774981611,"id":47591492,"options":[],"parent_id":47585636,"points":null,"story_id":47584540,"text":"How so? Good bit of my global claude.md is dedicated to fighting the incessant attribution in git commits. It is on the same level as the "sent from my iphone" signature - I'm not okay with my commits being advertising board for anthropic.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:11:48.000Z","created_at_i":1774955508,"id":47585636,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"That whole \u201cfeature\u201d is vile.","title":null,"type":"comment","url":null},{"author":"mrlnstk","children":[{"author":"BoppreH","children":[{"author":"bhaak","children":[],"created_at":"2026-03-31T13:08:02.000Z","created_at_i":1774962482,"id":47586823,"options":[],"parent_id":47585690,"points":null,"story_id":47584540,"text":"A benign use of this mode is developing on their own public repositories.
https://github.com/anthropics/claude-code","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:16:29.000Z","created_at_i":1774955789,"id":47585690,"options":[],"parent_id":47585675,"points":null,"story_id":47584540,"text":"I don't care who is using it, I don't want LLMs pretending to be humans in public repos. Anthropic just lost some points with me for this one.
EDIT: I just realized this might be used without publishing the changes, for internal evaluation only as you mentioned. That would be a lot better.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:14:37.000Z","created_at_i":1774955677,"id":47585675,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"But will this be released as a feature? For me it seems like it's an Anthropic internal tool to secretly contribute to public repositories to test new models etc.","title":null,"type":"comment","url":null},{"author":"0x3f","children":[{"author":"RockRobotRock","children":[{"author":"monsieurbanana","children":[],"created_at":"2026-03-31T12:27:06.000Z","created_at_i":1774960026,"id":47586331,"options":[],"parent_id":47586168,"points":null,"story_id":47584540,"text":"That one is magical for sure
https://en.wikipedia.org/wiki/Magic_(illusion)","title":null,"type":"comment","url":null},{"author":"TrickyRick","children":[],"created_at":"2026-03-31T12:44:00.000Z","created_at_i":1774961040,"id":47586538,"options":[],"parent_id":47586168,"points":null,"story_id":47584540,"text":"Giving your retina scan to one of the main Slop Bros, what could possibly go wrong?","title":null,"type":"comment","url":null},{"author":"stackghost","children":[],"created_at":"2026-03-31T17:11:13.000Z","created_at_i":1774977073,"id":47590455,"options":[],"parent_id":47586168,"points":null,"story_id":47584540,"text":"Scam Altman is not trustworthy. I hope nobody gives him their biometrics. I certainly would never.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:10:45.000Z","created_at_i":1774959045,"id":47586168,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":">There's no magical human verification technology coming to save us.
Except for the one Sam Altman is building.","title":null,"type":"comment","url":null},{"author":"matkoniecz","children":[{"author":"nslsm","children":[{"author":"primevaldad","children":[],"created_at":"2026-03-31T17:23:27.000Z","created_at_i":1774977807,"id":47590643,"options":[],"parent_id":47587848,"points":null,"story_id":47584540,"text":"and naming your feelings is the first step toward restoration","title":null,"type":"comment","url":null},{"author":"matkoniecz","children":[],"created_at":"2026-03-31T18:58:48.000Z","created_at_i":1774983528,"id":47591889,"options":[],"parent_id":47587848,"points":null,"story_id":47584540,"text":"Probably beats being in denial over it and pretending you like it.
And identifying problem you dislike is a good first step to find a strategy to solve it at least in part.","title":null,"type":"comment","url":null},{"author":"JoshTriplett","children":[{"author":"nslsm","children":[{"author":"JoshTriplett","children":[{"author":"nslsm","children":[{"author":"JoshTriplett","children":[{"author":"nslsm","children":[],"created_at":"2026-03-31T20:28:48.000Z","created_at_i":1774988928,"id":47593030,"options":[],"parent_id":47592965,"points":null,"story_id":47584540,"text":"Before you decide not to care about something, you are supposed to make a deep assessment to see whether you can change it. It is only after you\u2019ve determined that the thing can\u2019t be changed that you can choose not to care about it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:23:25.000Z","created_at_i":1774988605,"id":47592965,"options":[],"parent_id":47592932,"points":null,"story_id":47584540,"text":"You said "can\u2019t change". I observed that deciding you can't change something is self-fulfilling. Your argument from that point still relied on the assumption that you can't change it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:21:09.000Z","created_at_i":1774988469,"id":47592932,"options":[],"parent_id":47592820,"points":null,"story_id":47584540,"text":"That\u2019s a reductionist and wrong reading of the argument I made.","title":null,"type":"comment","url":null},{"author":"0x3f","children":[],"created_at":"2026-04-01T08:16:57.000Z","created_at_i":1775031417,"id":47598243,"options":[],"parent_id":47592820,"points":null,"story_id":47584540,"text":"You can care about things, but it seems preferable to care about that which you can change","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:13:12.000Z","created_at_i":1774987992,"id":47592820,"options":[],"parent_id":47592766,"points":null,"story_id":47584540,"text":""It's uncool to care about things" is, fortunately, not a compelling argument for people who care about things.
This tangent does not seem likely to go anywhere productive.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:08:34.000Z","created_at_i":1774987714,"id":47592766,"options":[],"parent_id":47592604,"points":null,"story_id":47584540,"text":"Which is not a problem if you choose not to worry about it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:55:53.000Z","created_at_i":1774986953,"id":47592604,"options":[],"parent_id":47587848,"points":null,"story_id":47584540,"text":"Deciding that you can't change something is the first and last step towards failing to change it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:24:20.000Z","created_at_i":1774967060,"id":47587848,"options":[],"parent_id":47586252,"points":null,"story_id":47584540,"text":"Feeling bad about something you can\u2019t change is bad for your mental health.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:19:26.000Z","created_at_i":1774959566,"id":47586252,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Even if it is impossible to win, I am still feeling bad about it.
And at this point it is more about how large space will be usable and how much will be bot-controlled wasteland. I prefer spaces important for me to survive.","title":null,"type":"comment","url":null},{"author":"xyzal","children":[{"author":"satvikpendem","children":[{"author":"paradox460","children":[],"created_at":"2026-03-31T14:42:18.000Z","created_at_i":1774968138,"id":47588073,"options":[],"parent_id":47587344,"points":null,"story_id":47584540,"text":"I used to leave a megahal connected to my bouncer when I wasn't around","title":null,"type":"comment","url":null},{"author":"al_borland","children":[{"author":"satvikpendem","children":[],"created_at":"2026-03-31T18:01:09.000Z","created_at_i":1774980069,"id":47591181,"options":[],"parent_id":47591084,"points":null,"story_id":47584540,"text":"It was for fun, to see if it were possible and whether others could detect they were talking to a bot or not, you know, the hacker ethos and all. It's not meant to be taken seriously although looks like these days people unironically have LLM "friends."","title":null,"type":"comment","url":null},{"author":"xyzal","children":[],"created_at":"2026-04-01T07:41:15.000Z","created_at_i":1775029275,"id":47598015,"options":[],"parent_id":47591084,"points":null,"story_id":47584540,"text":"Exactly. Pick friends that do not behave like dicks.","title":null,"type":"comment","url":null},{"author":"0x3f","children":[],"created_at":"2026-04-01T08:23:10.000Z","created_at_i":1775031790,"id":47598287,"options":[],"parent_id":47591084,"points":null,"story_id":47584540,"text":"I think you underestimate how many people already run their opinions and responses through LLMs, even if the LLM is not writing them wholesale. Intelligence is part of the social game, so appearing to have it matters to people. Friend groups are just social groups of a certain kind, they're not really removed from all this.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:53:29.000Z","created_at_i":1774979609,"id":47591084,"options":[],"parent_id":47587344,"points":null,"story_id":47584540,"text":"I don't see what the value of this would be. Why would I want to automate talking to my friends? If I'm not interested in talking with them, I could simply not do it. It also carries the risk of not actually knowing what was talked about or said, which could come up in real life and lead to issues. If a "friend" started using a bot to talk to me, they would not longer be considered a friend. That would be the end.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:46:46.000Z","created_at_i":1774964806,"id":47587344,"options":[],"parent_id":47587194,"points":null,"story_id":47584540,"text":"Until your friend writes a bot.
Funny story, when I was younger I trained a basic text predictor deep learning model on all my conversations in a group chat I was in, it was surprisingly good at sounding like me and sometimes I'd use it to generate some text to submit to the chat.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:37:06.000Z","created_at_i":1774964226,"id":47587194,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Magical human verification technology is called "your own private forum" in conjunction with "invite your friends"","title":null,"type":"comment","url":null},{"author":"keybored","children":[],"created_at":"2026-03-31T13:40:23.000Z","created_at_i":1774964423,"id":47587246,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Negative sentiment towards technological destiny detected in human agent.","title":null,"type":"comment","url":null},{"author":"jesse_dot_id","children":[{"author":"mr_00ff00","children":[],"created_at":"2026-03-31T15:37:24.000Z","created_at_i":1774971444,"id":47588968,"options":[],"parent_id":47588518,"points":null,"story_id":47584540,"text":"That\u2019s like having your security on the frontend.
If someone owns the keyboard then they can fake those metrics and tell the server it is happening when it isn\u2019t.
That will be easy to beat.","title":null,"type":"comment","url":null},{"author":"0x3f","children":[],"created_at":"2026-04-01T08:19:14.000Z","created_at_i":1775031554,"id":47598262,"options":[],"parent_id":47588518,"points":null,"story_id":47584540,"text":"We already do this, and with mouse. It's been defeated long ago.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:11:17.000Z","created_at_i":1774969877,"id":47588518,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"I assume we're heading to a place where keyboards will all have biometric sensors on every key and measure weight fluctuations in keystrokes, actually.","title":null,"type":"comment","url":null},{"author":"j2kun","children":[{"author":"WarmWash","children":[],"created_at":"2026-03-31T17:28:42.000Z","created_at_i":1774978122,"id":47590731,"options":[],"parent_id":47589357,"points":null,"story_id":47584540,"text":"I think "adapt or die" is the takeaway.","title":null,"type":"comment","url":null},{"author":"MidnightRider39","children":[{"author":"DaiPlusPlus","children":[{"author":"lazyasciiart","children":[],"created_at":"2026-03-31T19:13:53.000Z","created_at_i":1774984433,"id":47592079,"options":[],"parent_id":47591850,"points":null,"story_id":47584540,"text":"I got hit by a car while out for a run. Just saying.","title":null,"type":"comment","url":null},{"author":"MidnightRider39","children":[],"created_at":"2026-03-31T21:54:28.000Z","created_at_i":1774994068,"id":47593987,"options":[],"parent_id":47591850,"points":null,"story_id":47584540,"text":"You can probably find a million situations where doing less is terrible.
I think first step would be to define for yourself what doing less actually means - it could mean taking a walk instead of chasing dopamine -> doing less but you move more.
But whatever it\u2019s a philosophical question and there aren\u2019t any right or true answers","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:56:24.000Z","created_at_i":1774983384,"id":47591850,"options":[],"parent_id":47591082,"points":null,"story_id":47584540,"text":"> I think generally doing less is better.
My sedentary lifestyle is responsible for my recurrent cellulitis infections.
Just saying.","title":null,"type":"comment","url":null},{"author":"skybrian","children":[{"author":"MidnightRider39","children":[],"created_at":"2026-03-31T21:52:25.000Z","created_at_i":1774993945,"id":47593970,"options":[],"parent_id":47592385,"points":null,"story_id":47584540,"text":"I didn\u2019t mean that you should strive to do as little as possible; rather that if you have 2 choices, do more or do less - then I would be biased towards doing less. \nOf course not always a realistic option","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:39:51.000Z","created_at_i":1774985991,"id":47592385,"options":[],"parent_id":47591082,"points":null,"story_id":47584540,"text":"On the other hand, if a dead person can do it better than you can, it's not that much of an accomplishment.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:53:23.000Z","created_at_i":1774979603,"id":47591082,"options":[],"parent_id":47589357,"points":null,"story_id":47584540,"text":"This is the only relevant question. And it leads right to the next one which is \u201cwhat is a good life?\u201d
But humans have a huge bias for action. I think generally doing less is better.","title":null,"type":"comment","url":null},{"author":"tombert","children":[{"author":"palmotea","children":[],"created_at":"2026-03-31T18:13:36.000Z","created_at_i":1774980816,"id":47591344,"options":[],"parent_id":47591238,"points":null,"story_id":47584540,"text":"> I feel that fatalism, especially when people treat it as some sort of personal philosophy, is kind of lazy.
I think a lot of fatalism is fake. It's really someone saying "I like this, and I want you to believe you can't change it so you give up."","title":null,"type":"comment","url":null},{"author":"Nevermark","children":[],"created_at":"2026-03-31T18:21:11.000Z","created_at_i":1774981271,"id":47591440,"options":[],"parent_id":47591238,"points":null,"story_id":47584540,"text":"Apathy has a striking number of motivated evangelists!","title":null,"type":"comment","url":null},{"author":"gspr","children":[{"author":"joquarky","children":[{"author":"tombert","children":[{"author":"ykonstant","children":[{"author":"gspr","children":[{"author":"ykonstant","children":[{"author":"tombert","children":[],"created_at":"2026-04-01T17:43:34.000Z","created_at_i":1775065414,"id":47604086,"options":[],"parent_id":47598351,"points":null,"story_id":47584540,"text":"You are kind of proving my point. You are actively justifying doing literally nothing about what bothers you and acting indignant and self righteous about it.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T08:32:17.000Z","created_at_i":1775032337,"id":47598351,"options":[],"parent_id":47597916,"points":null,"story_id":47584540,"text":"Don't spread the misery?? Wow, fucking thanks.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T07:24:28.000Z","created_at_i":1775028268,"id":47597916,"options":[],"parent_id":47597487,"points":null,"story_id":47584540,"text":"> Telling random citizens "run for office" is facetious, whether you mean it or not.
Hard disagree. At least where I live, "random citizens" run for local office and succeed all the time.
Also, complaining that you "have no network" is a you problem, not a system problem. I'm truly sorry if you feel you have no friends, but you'll be better off at least trying to get some (independent of politics). And if that's something you've tried and failed at before, I do feel pity. But I don't think hope is lost for anyone. And even if it were lost, please don't actively spread the misery!","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T06:20:06.000Z","created_at_i":1775024406,"id":47597487,"options":[],"parent_id":47592911,"points":null,"story_id":47584540,"text":">Vote for better politicians.
I cannot. I can only vote better politicians if they are there. That is without even going into the minefield of what is "better". My implication is that I have no confidence whatsoever in any current politician in my state.
> Donate money to causes that you think are important.
I have no money.
> If you think you can do it better, and this isn't meant to be facetious, run for political office.
I have no money, no visibility and no connections. Even if I was magically given tons of money, I would still need a strong network to attempt any real change, even without taking into consideration the strong networks already in place preventing it.
Telling random citizens "run for office" is facetious, whether you mean it or not.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:19:45.000Z","created_at_i":1774988385,"id":47592911,"options":[],"parent_id":47592582,"points":null,"story_id":47584540,"text":"Then do something about it. Vote for better politicians. Donate money to causes that you think are important. If you think you can do it better, and this isn't meant to be facetious, run for political office.
Being fatalistic can be a great excuse not to do anything.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:54:27.000Z","created_at_i":1774986867,"id":47592582,"options":[],"parent_id":47592077,"points":null,"story_id":47584540,"text":"Perhaps the current societal trajectory is destroying everything that they hold dear.
I mean, just look around you.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:13:49.000Z","created_at_i":1774984429,"id":47592077,"options":[],"parent_id":47591238,"points":null,"story_id":47584540,"text":"It also makes no sense! "Fuck this, it doesn't matter - but I'll happily spend effort communicating that to others, because apparently making others not care about something I don't care about is something I do care about." Wut?!
Well, I say it makes no sense. Alternatively, it makes a lot of sense, and these people actually just wanna destroy everything we hold dear :-(","title":null,"type":"comment","url":null},{"author":"joquarky","children":[],"created_at":"2026-03-31T19:52:24.000Z","created_at_i":1774986744,"id":47592550,"options":[],"parent_id":47591238,"points":null,"story_id":47584540,"text":"This is more cultural rather than rational.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:04:59.000Z","created_at_i":1774980299,"id":47591238,"options":[],"parent_id":47589357,"points":null,"story_id":47584540,"text":"I feel that fatalism, especially when people treat it as some sort of personal philosophy, is kind of lazy.
It requires no effort to say "fuck this, nothing matters anyway", and then justify doing literally nothing.","title":null,"type":"comment","url":null},{"author":"mikkupikku","children":[{"author":"JoshTriplett","children":[{"author":"mikkupikku","children":[],"created_at":"2026-03-31T20:09:21.000Z","created_at_i":1774987761,"id":47592778,"options":[],"parent_id":47592571,"points":null,"story_id":47584540,"text":"Cringe.","title":null,"type":"comment","url":null},{"author":"bandrami","children":[{"author":"JoshTriplett","children":[{"author":"bandrami","children":[],"created_at":"2026-04-01T02:20:44.000Z","created_at_i":1775010044,"id":47596012,"options":[],"parent_id":47595675,"points":null,"story_id":47584540,"text":"The difference between "essentially infinite" and "actually infinite". Infinity is a very long time.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T01:27:47.000Z","created_at_i":1775006867,"id":47595675,"options":[],"parent_id":47592991,"points":null,"story_id":47584540,"text":"I'm sorry to hear that you don't want to exist in the future. I do. I have thought about it extensively, and there is literally no scenario in which I consider not-existing better than existing.
There is an essentially infinite amount of creativity and interesting complexity available in the richness of interactions with other people and the things people create. What, exactly, are you "horrified" about?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:25:57.000Z","created_at_i":1774988757,"id":47592991,"options":[],"parent_id":47592571,"points":null,"story_id":47584540,"text":"I can think of no concept more horrifying than personal immortality and if you disagree I don't think you've thought about it enough.","title":null,"type":"comment","url":null},{"author":"jacquesm","children":[{"author":"JoshTriplett","children":[],"created_at":"2026-04-01T01:25:51.000Z","created_at_i":1775006751,"id":47595666,"options":[],"parent_id":47593139,"points":null,"story_id":47584540,"text":"s/make peace with it/make war with it/. To the last breath.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:35:46.000Z","created_at_i":1774989346,"id":47593139,"options":[],"parent_id":47592571,"points":null,"story_id":47584540,"text":"No, there is no alternative. Everything eventually dies, so you better make peace with it. The only people who believe that they won't die are religious people who believe in an afterlife (which is a preposterous position) and the people who have their heads or whole bodies frozen because they think they are so special that the future will honor their contracts and revive them.
Both of these are bound to lead to the exact same outcome so it doesn't really matter what you believe but it may guide you to wiser decision while you are alive to accept reality absent proof to the contrary.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:53:50.000Z","created_at_i":1774986830,"id":47592571,"options":[],"parent_id":47591307,"points":null,"story_id":47584540,"text":"> Do you want to live forever?
Yes, of course. Do you prefer to die? Those are the only two alternatives, and a decision that you don't want one is a decision that you prefer the other.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:11:16.000Z","created_at_i":1774980676,"id":47591307,"options":[],"parent_id":47589357,"points":null,"story_id":47584540,"text":"> But I suppose death comes for us all, yes? Why do anything at all?
Wrong take. Death comes for us all, yes, so why hold back? Do you want to live forever?","title":null,"type":"comment","url":null},{"author":"0x3f","children":[],"created_at":"2026-04-01T07:45:22.000Z","created_at_i":1775029522,"id":47598039,"options":[],"parent_id":47589357,"points":null,"story_id":47584540,"text":"It's fun to pet the cat. It's not fun to rage against an unstoppable force. Well, maybe it is for some people. But I find people often underestimate the detrimental effects.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:59:47.000Z","created_at_i":1774972787,"id":47589357,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Fatalism will also not fix anything. But I suppose death comes for us all, yes? Why do anything at all?","title":null,"type":"comment","url":null},{"author":"stackghost","children":[{"author":"tlonny","children":[{"author":"stackghost","children":[],"created_at":"2026-03-31T17:05:01.000Z","created_at_i":1774976701,"id":47590368,"options":[],"parent_id":47590279,"points":null,"story_id":47584540,"text":"I'm hoping to do a Show HN soon :)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:58:49.000Z","created_at_i":1774976329,"id":47590279,"options":[],"parent_id":47590026,"points":null,"story_id":47584540,"text":"how does it work?","title":null,"type":"comment","url":null},{"author":"anonym29","children":[{"author":"stackghost","children":[],"created_at":"2026-03-31T17:06:02.000Z","created_at_i":1774976762,"id":47590383,"options":[],"parent_id":47590319,"points":null,"story_id":47584540,"text":"This is true, but I think there is a sizable (and growing) appetite for human-only spaces.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:01:43.000Z","created_at_i":1774976503,"id":47590319,"options":[],"parent_id":47590026,"points":null,"story_id":47584540,"text":"The technical implementation is the easy part. The hard part is achieving mass voluntary cooperation under adverse incentive schemes.","title":null,"type":"comment","url":null},{"author":"lynx97","children":[{"author":"stackghost","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T18:06:50.000Z","created_at_i":1774980410,"id":47591257,"options":[],"parent_id":47590445,"points":null,"story_id":47584540,"text":"Not parent poster but I am a maintainer of software powering significant portions of the internet and prove my humanity with a 16 year old PGP key with thousands of transitive trust signatures formed through mostly in-person meetings, using IETF standards and keychain smartcards, as is the case for everyone I work with.
But, I do not have an Android or iOS device as I do not use proprietary software, so a smartphone based solution would not work for me.
Why re-invent the wheel? Invest in making PGP easier and keep the decades of trust building going anchoring humans to a web of trust that long predates human-impersonation-capable AI.","title":null,"type":"comment","url":null},{"author":"lynx97","children":[],"created_at":"2026-04-01T12:00:41.000Z","created_at_i":1775044841,"id":47599676,"options":[],"parent_id":47590445,"points":null,"story_id":47584540,"text":"So, you haven't thought about it yet? Your counterquestion is insufficient to get any further, because "use" is a very relative term. If I can "use" a smartphone depends on the way you code your app. If you use inherently inaccessible UI elements, I can't "use" your app, no matter if I own a smartphone or not.
I might reply with a similarily useless question: Can you write accessible smartphone apps?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:10:14.000Z","created_at_i":1774977014,"id":47590445,"options":[],"parent_id":47590413,"points":null,"story_id":47584540,"text":"I think so? Can you use a smartphone?
edit: can't reply, the rate-limiting is such an awful UX","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:08:12.000Z","created_at_i":1774976892,"id":47590413,"options":[],"parent_id":47590026,"points":null,"story_id":47584540,"text":"Great! Lets do the CAPTCHA-test: Will I, as a 100% blind user, be able to complete your process?","title":null,"type":"comment","url":null},{"author":"lrvick","children":[{"author":"sebastiennight","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T22:35:31.000Z","created_at_i":1774996531,"id":47594368,"options":[],"parent_id":47591308,"points":null,"story_id":47584540,"text":"They need to go to generate their key, ideally offline with an offline CA backup on and subkeys on a nitrokey or yubikey smartcard with touch requirement enabled for all key operations for safe workstation use. One can use keyfork on AirgapOS to do this safely, as a once-ever operation.
From there they set up their workstation tools to sign every ssh connection, git push, commit, merge, review, secret decryption, and release signature with their PGP smartcard which is all very well supported. This offers massive damage control if you get malware on their system, in addition to preventing online impersonation.
From there they ideally link it to all their online accounts with keyoxide to make it easy to verify as a single long lived identity, then start seeking out key signing parties locally or at tech conferences, hackerspaces etc.
We run one at CCC most years at the Church Of Cryptography.
Think of it like a long term digital passport that requires a few signatures by an international set of human notarys before anyone significantly trusts it.
Yes it requires a manual set of human steps anchored to human reputation online and offline, which is a doorway swarms of made up AI bot identities cannot pass through.
Do I expect most humans to do this? Absolutely not. However I consider it _negligent_ for any maintainer of a widely used open source software project to _not_ do this or they risk an impersonator pushing malware to their users.
No idea on theoretical rate of expansion but all the major security conscious classic linux distros mandate this for all maintainers. There are only maybe 20k people on earth that significantly contribute to FOSS internet foundations and Linux distros, so it scales just fine there.
Note: with the exception of stagex, most modern distros like alpine and nix have a yolo wikipedia style trust model, so never ever use those in production.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:11:20.000Z","created_at_i":1774980680,"id":47591308,"options":[],"parent_id":47591193,"points":null,"story_id":47584540,"text":"Say a regular human wanted to join and prove their humanhood status (expanding the web of trust). How would they go about that? What is the theoretical ceiling on the rate of expansion of this implementation?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:02:00.000Z","created_at_i":1774980120,"id":47591193,"options":[],"parent_id":47590026,"points":null,"story_id":47584540,"text":"We already have it and we use it to validate the trusted human maintainer involvement behind the linux packages that power the entire internet: PGP Web Of Trust. Still works as designed and I still go to keysigning parties in person.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:41:07.000Z","created_at_i":1774975267,"id":47590026,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"I am actively building non-magical human verification technology that doesn't require you uploading your retinal scans or ID to billionaires or incompetent outsourcing firms.","title":null,"type":"comment","url":null},{"author":"SV_BubbleTime","children":[{"author":"saltcured","children":[],"created_at":"2026-03-31T18:03:23.000Z","created_at_i":1774980203,"id":47591215,"options":[],"parent_id":47590857,"points":null,"story_id":47584540,"text":"How quickly would that meat-space renaissance spin through our whole cyberpunk heritage, speedrunning the same authentication challenges..?
The cornucopia of gargoyles, living their best life as terminals for the machine.
The strange p-zombies who don't show their gargoyle accessories visibly, but somehow still follow the script.
Eventually the more insidious infiltrators, requiring a real Voight-Kampff test.","title":null,"type":"comment","url":null},{"author":"sebastiennight","children":[],"created_at":"2026-03-31T18:08:33.000Z","created_at_i":1774980513,"id":47591277,"options":[],"parent_id":47590857,"points":null,"story_id":47584540,"text":""minority"?
This is https://en.wikipedia.org/wiki/Sybil_attack","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:36:44.000Z","created_at_i":1774978604,"id":47590857,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"IDK. I sort of like the idea that now instead of dead internet theory being a joke, that it\u2019ll be a well known fact that a minority of people are not real and there is no point in engaging\u2026 I look forward to Social 3\u2026 where people have to meet face to face.","title":null,"type":"comment","url":null},{"author":"ex-aws-dude","children":[],"created_at":"2026-03-31T17:43:42.000Z","created_at_i":1774979022,"id":47590956,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"That's why I stopped brushing my teeth, I can't clean every crevice perfectly so what's the point?","title":null,"type":"comment","url":null},{"author":"lrvick","children":[{"author":"srmatto","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T21:30:01.000Z","created_at_i":1774992601,"id":47593703,"options":[],"parent_id":47591298,"points":null,"story_id":47584540,"text":"The people who signed my keys trust me to be an honest human actor that chose this as the singular identity they signed for the human body they met in person.
I -could- burn my 16+ years of reputation by letting a bot start signing commits as me, and I could also set my house on fire. I have very strong incentive not to do so as my aggregate trust is very expensive and the humans that signed me would be unlikely to sign a second if I ruined the reputation of my first.
This incentive structure is why web of trust actually works pretty well, and is the best "proof of human" we are likely ever going to have while respecting privacy and anonymity for those that need it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:10:50.000Z","created_at_i":1774980650,"id":47591298,"options":[],"parent_id":47591145,"points":null,"story_id":47584540,"text":"Can't you just instruct Claude Code to use your signing keys? I understand you may say "I won't." But my point is that someone can.","title":null,"type":"comment","url":null},{"author":"nothrabannosir","children":[{"author":"lrvick","children":[{"author":"nothrabannosir","children":[{"author":"lrvick","children":[{"author":"nothrabannosir","children":[],"created_at":"2026-03-31T23:14:09.000Z","created_at_i":1774998849,"id":47594715,"options":[],"parent_id":47594588,"points":null,"story_id":47584540,"text":"Your last point, I think, is why so many sibling commenters are balking at GP :)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:59:05.000Z","created_at_i":1774997945,"id":47594588,"options":[],"parent_id":47594486,"points":null,"story_id":47584540,"text":"I suppose you are correct. I am agreeing that if one widely deployed the defense tactics projects like stagex use, then asshats using things like undercover will not be trusted.
Unfortunately outside of classic Linux packaging platforms, useful web of trust and signing is very rare, so I expect things like undercover mode being popular are going to make everything a lot worse before it gets better.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:48:02.000Z","created_at_i":1774997282,"id":47594486,"options":[],"parent_id":47593883,"points":null,"story_id":47584540,"text":"Exactly. So in the words of the comment you replied to: why are we wasting energy on worrying about Claude code impersonating humans? We have that solution you proposed.
That\u2019s what I mean by \u201cyou agree with the person to whom you replied\u201d","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:44:41.000Z","created_at_i":1774993481,"id":47593883,"options":[],"parent_id":47591393,"points":null,"story_id":47584540,"text":"I can prove that code was signed by a key that was verified to belong to a single human body by lots of in-person high reputation humans.
How the code was authored, who cares, but I can prove it had multiple explicit cryptographic human signoffs before merge, and that is what matters in terms of quality control and supply chain attack resistance.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:17:47.000Z","created_at_i":1774981067,"id":47591393,"options":[],"parent_id":47591145,"points":null,"story_id":47584540,"text":"You can prove the commits were signed by a key you once verified. It is your trust in those people which allows you to extend that to \u201cno LLM\u201d usage, but that\u2019s reframing the conversation as one of trust, not human / machine. Which is (charitably) GPs point: stop framing this as machine vs human \u2014 assume (\u201caccept\u201d) that all text can be produced by machines and go from there: what now? That\u2019s where your proposal is one solution: strict web of trust. It has pros and cons (barrier to entry for legitimate first timers), but it\u2019s a valid proposal.
All that to say \u201cyou\u2019re not disagreeing with the person you\u2019re replying to\u201d lol xD","title":null,"type":"comment","url":null},{"author":"orf","children":[{"author":"lrvick","children":[{"author":"orf","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T22:48:30.000Z","created_at_i":1774997310,"id":47594494,"options":[],"parent_id":47593983,"points":null,"story_id":47584540,"text":"It is not a silver bullet by itself, but when combined with the other tactics in stagex I believe it gives us a very strong supply chain attack defense.
I can not prove the tools used, but I can prove multiple humans signed off on code with keys they stake their personal reputations on that I have confirmed they maintain on smartcards.
While nothing involving humans is perfect I feel it is best effort with existing tools and standards and makes us one of the hardest projects to deploy a successful supply chain attack on today.
Edit: Saw your edit. Thanks!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:54:00.000Z","created_at_i":1774994040,"id":47593983,"options":[],"parent_id":47593745,"points":null,"story_id":47584540,"text":"That\u2019s great, and I also don\u2019t care. But I think all people are saying is that by most definitions you cannot \u201cprove all contributions to stagex are by humans\u201d.
Or are you saying you can prove that aliens and cats didn\u2019t make them? Because I\u2019m not sure that\u2019s true either.
And once you find out someone has trained their dog to commit something, how exactly do you revoke your trust?
I think if you answer these questions you\u2019ll see pretty quickly why this solution isn\u2019t the silver bullet you think it is.
Edit: stagex looks really, really good","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:33:23.000Z","created_at_i":1774992803,"id":47593745,"options":[],"parent_id":47591605,"points":null,"story_id":47584540,"text":"I absolutely do not care what autocomplete tools someone used. Only that they as humans own and sign what is submitted so it is attached to their very expensive reputations they do not want to lose.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:36:34.000Z","created_at_i":1774982194,"id":47591605,"options":[],"parent_id":47591145,"points":null,"story_id":47584540,"text":"You can only prove that all contributions are pushed by those humans, and you can quite explicitly/clearly not prove that those humans didn't use any AI prior to pushing.","title":null,"type":"comment","url":null},{"author":"spullara","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T21:36:54.000Z","created_at_i":1774993014,"id":47593784,"options":[],"parent_id":47591863,"points":null,"story_id":47584540,"text":"Do you think it is likely that the majority of the people that spent decades building this trust graph and gaining the trust needed to be release engineers on the packages that power the whole internet are just going to hand off control of that key to a bot?
Anyone doing so would be setting their professional reputations completely on fire, and burning your in-person-built web of trust is a once in a lifetime thing.
Basically, we trust the keys belong to humans and are controlled by humans because to do otherwise would be a violation of the universally understood trust contract and would thus be reputational bankruptcy that would take years to overcome, if ever.
Even so, we assume at least one maintainer is dishonest at all times, which is why every action needs signatures from two or more maintainers.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:57:07.000Z","created_at_i":1774983427,"id":47591863,"options":[],"parent_id":47591145,"points":null,"story_id":47584540,"text":"nothing about this proves anything except that someone or something had access to the key.","title":null,"type":"comment","url":null},{"author":"jacquesm","children":[{"author":"lrvick","children":[{"author":"jacquesm","children":[{"author":"lrvick","children":[{"author":"jacquesm","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T22:53:49.000Z","created_at_i":1774997629,"id":47594548,"options":[],"parent_id":47594072,"points":null,"story_id":47584540,"text":"I am reasonably confident it is the current industry best effort, and way beyond the status quo, not that it is perfect.
We combine many tactics for defense in depth that I strongly suspect if widely deployed would put a stop to the daily supply chain attack headlines.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:03:27.000Z","created_at_i":1774994607,"id":47594072,"options":[],"parent_id":47593831,"points":null,"story_id":47584540,"text":"While I appreciate all of the effort you put in this and respect that you trust this to be bulletproof I'm always going to be skeptical of silver bullets.
Your level of certainty is the thing that frightens me more than the confidence I have in the quality of your work.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:40:44.000Z","created_at_i":1774993244,"id":47593831,"options":[],"parent_id":47593787,"points":null,"story_id":47584540,"text":"If we were talking about any linux distribution before stagex, I would agree with you.
Stagex however expects at least one maintainer may at any time engage in reputation-ending dishonesty or simply they were threatened or coerced. This is why every single release is signed by a -quorum- of code reviewers and code reproducers that must all build locally and get identical hashes, so no single points of failure exist in our trust graph.
Our last release was signed by four geodistributed maintainers that all attest to having built the entire distribution from 180 bytes of machine code all the way up with the same hashes.
All of their keys being compromised at once gets beyond the pale.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:37:11.000Z","created_at_i":1774993031,"id":47593787,"options":[],"parent_id":47593721,"points":null,"story_id":47584540,"text":"The problem is all it really takes is one edge case to successfully break a web of trust to the point that the web of trust becomes a blind spot. Instead of distrusting everybody (which should be the default) the web of trust attempts to create a 'walled garden of trust' and behind that wall everybody can be friendly. That gives a successful attacker a massive advantage.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:31:56.000Z","created_at_i":1774992716,"id":47593721,"options":[],"parent_id":47592411,"points":null,"story_id":47584540,"text":"Of course! There are always edge cases, but I would suspect the number of bots signed by reputable keys to be near 0%, and the honest human score in this trust graph to be well over 90%.
Compare to how much we should trust any random unsigned key signing commits, or unsigned commits, in which the trust should be 0% unless you have reviewed the code yourself.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:42:19.000Z","created_at_i":1774986139,"id":47592411,"options":[],"parent_id":47591145,"points":null,"story_id":47584540,"text":"With 5400+ people I am betting that you have at least one person in your 'web of trust' that no longer deserves that trust.
That's one of the intrinsic problems with webs of trust (and with democracy...), you extend your trust but it does not automatically revoke when the person can no longer be trusted.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:57:35.000Z","created_at_i":1774979855,"id":47591145,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"I can prove all contributions to stagex are by humans because we all belong to a 25 year old web of trust with 5444 endorser keys including most redhat, debian, ubuntu, and fedora maintainers, with all of our own maintainer keys in smartcards we tap to sign every review and commit, and we do background checks on every new maintainer.
I am completely serious. We have always had a working proof of human system called Web of Trust and while everyone loves to hate on PGP (in spite of it using modern ECC crypto these days) it is the only widely deployed spec that solves this problem.
https://kron.fi/en/posts/stagex-web-of-trust/","title":null,"type":"comment","url":null},{"author":"thih9","children":[],"created_at":"2026-03-31T18:09:39.000Z","created_at_i":1774980579,"id":47591287,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Human verification technology absolutely exists. Give it some time and people who sell ai today are going to shoehorn it everywhere as the solution to the problem they are busy creating now.","title":null,"type":"comment","url":null},{"author":"taurath","children":[],"created_at":"2026-03-31T18:10:21.000Z","created_at_i":1774980621,"id":47591293,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"It\u2019s where THIS internet is headed. The future may involve a lot more of them I think.","title":null,"type":"comment","url":null},{"author":"marricks","children":[],"created_at":"2026-03-31T18:30:34.000Z","created_at_i":1774981834,"id":47591528,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Nothing like throwing in the towel before a battle is ever fought. Let's just sigh and wearily march on to our world of AI slop and ever higher bug counts and latency delays while we wait for the five different phone homes and compilations through a billion different LLM's for every silly command.","title":null,"type":"comment","url":null},{"author":"themafia","children":[{"author":"joquarky","children":[{"author":"themafia","children":[],"created_at":"2026-03-31T23:51:01.000Z","created_at_i":1775001061,"id":47595009,"options":[],"parent_id":47592683,"points":null,"story_id":47584540,"text":"Hardly. They're clearly trying to influence a group at large using a flawed logical tactic. I'm pointing that out with a device that suggests their same futility they expect others to adapt should be primarily experienced by them instead.
If pointing out bullies is bullying then you're in a ridiculous mindset.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:02:01.000Z","created_at_i":1774987321,"id":47592683,"options":[],"parent_id":47592013,"points":null,"story_id":47584540,"text":"That's an ironic comment.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:08:41.000Z","created_at_i":1774984121,"id":47592013,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"> You'll never win this battle, so why waste feelings and energy on it?
Cool. The attitude of a bully. Thanks for the contribution!","title":null,"type":"comment","url":null},{"author":"layer8","children":[],"created_at":"2026-03-31T19:08:52.000Z","created_at_i":1774984132,"id":47592014,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"Technology won\u2019t save us, but that doesn\u2019t mean we shouldn\u2019t be promoting ethics.","title":null,"type":"comment","url":null},{"author":"gspr","children":[{"author":"0x3f","children":[{"author":"gspr","children":[{"author":"0x3f","children":[{"author":"gspr","children":[],"created_at":"2026-04-02T18:37:24.000Z","created_at_i":1775155044,"id":47618400,"options":[],"parent_id":47611660,"points":null,"story_id":47584540,"text":"No.","title":null,"type":"comment","url":null}],"created_at":"2026-04-02T08:41:13.000Z","created_at_i":1775119273,"id":47611660,"options":[],"parent_id":47604954,"points":null,"story_id":47584540,"text":"Wasting your life fighting things that can't be fought is functionally equivalent to dying sooner.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T18:53:51.000Z","created_at_i":1775069631,"id":47604954,"options":[],"parent_id":47598248,"points":null,"story_id":47584540,"text":"This is such an incredibly imbecilic comment.
Listen to this guy: "because you don't take the ultimate risk for what you believe in, you are dumb for suggesting you should do anything whatsoever".
Go away. The world doesn't need your dark resignation.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T08:17:28.000Z","created_at_i":1775031448,"id":47598248,"options":[],"parent_id":47592032,"points":null,"story_id":47584540,"text":"Well why not head for the front lines of Ukraine? Or Russia, depending on your preference.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:10:06.000Z","created_at_i":1774984206,"id":47592032,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"To feel something. To resist something bad. To stand for what is right.
Do those sentiments mean nothing to you?","title":null,"type":"comment","url":null},{"author":"danny_codes","children":[{"author":"skybrian","children":[],"created_at":"2026-03-31T19:46:17.000Z","created_at_i":1774986377,"id":47592456,"options":[],"parent_id":47592086,"points":null,"story_id":47584540,"text":"I think you're underestimating the difficulty, even for exact copies of text (which AI mostly isn't doing).
What sort of Orwellian anti-cheat system would prevent copy and paste from working? What sort of law would mandate that? There are elaborate systems preventing people from copying video but they still have an analog hole.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:14:10.000Z","created_at_i":1774984450,"id":47592086,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"It\u2019s certainly winnable with some legislative tweaks. These systems are all designed by humans, we can just change them.
Of course, we\u2019d need a significant change of direction in leadership, but it\u2019s happened many times before. French Revolution seems highly relevant","title":null,"type":"comment","url":null},{"author":"xantronix","children":[{"author":"jacquesm","children":[{"author":"lrvick","children":[],"created_at":"2026-03-31T21:46:09.000Z","created_at_i":1774993569,"id":47593901,"options":[],"parent_id":47593082,"points":null,"story_id":47584540,"text":"https://hashbang.sh - "underground" for over 20 years and still running strong.","title":null,"type":"comment","url":null},{"author":"xantronix","children":[],"created_at":"2026-04-01T00:59:33.000Z","created_at_i":1775005173,"id":47595492,"options":[],"parent_id":47593082,"points":null,"story_id":47584540,"text":"Amen brother, this one will be for me and all my homies.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:32:42.000Z","created_at_i":1774989162,"id":47593082,"options":[],"parent_id":47592135,"points":null,"story_id":47584540,"text":"Why stop at one? Make more such underground community supported internets. The more the merrier. Monoculture ends with death. The only question is how long it will take.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:18:47.000Z","created_at_i":1774984727,"id":47592135,"options":[],"parent_id":47585733,"points":null,"story_id":47584540,"text":"I guess I could just curl up into fetal position and watch the world go by. But that's no fun. Why not dream big and shoot for the moon with kooky goals like, say, having an underground, community-supported internet where things are falling less to shit?
Belief in inevitability is a choice (except for maybe dying, I guess).","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:20:58.000Z","created_at_i":1774956058,"id":47585733,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"You'll never win this battle, so why waste feelings and energy on it? That's where the internet is headed. There's no magical human verification technology coming to save us.","title":null,"type":"comment","url":null},{"author":"shaky-carrousel","children":[{"author":"kingstnap","children":[{"author":"briHass","children":[{"author":"mickdarling","children":[],"created_at":"2026-03-31T20:37:03.000Z","created_at_i":1774989423,"id":47593151,"options":[],"parent_id":47588814,"points":null,"story_id":47584540,"text":"Hey, LLM, take a look at these multiple hundred emails and docs in my docs folder from the last few years, before I started using AI, that I wrote personally. create a list of all of the idiosyncrasies that I have in my writing. Create a file to remember that. And then use that to write any new text that'll be published so it sounds like my authentic voice. Thank you.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:28:58.000Z","created_at_i":1774970938,"id":47588814,"options":[],"parent_id":47586380,"points":null,"story_id":47584540,"text":"Hey LLM, write me a system prompt that will avoid the common AI 'tells' or other idiosyncrasies that make it obvious that text or code output was generated by an AI/LLM. Use the referenced Wikipedia article as a must-avoid list, but do not consider it exhaustive. Add any derivations or modifications to these rules to catch 'likely' signals as well.
There, sorted!","title":null,"type":"comment","url":null},{"author":"skeledrew","children":[],"created_at":"2026-03-31T17:31:27.000Z","created_at_i":1774978287,"id":47590766,"options":[],"parent_id":47586380,"points":null,"story_id":47584540,"text":"All the prompts I've ever written with Claude have always worked fine the first time. Only revised if the actual purpose changes, I left something out, etc. But also I tend to only write prompts as part of a larger session, usually near the end, so there's lots of context available to help with the writing.","title":null,"type":"comment","url":null},{"author":"alwillis","children":[{"author":"joquarky","children":[],"created_at":"2026-03-31T20:11:06.000Z","created_at_i":1774987866,"id":47592796,"options":[],"parent_id":47591378,"points":null,"story_id":47584540,"text":"As someone who often fails to read subtext, I would estimate that most people expect you to participate in mind reading as a natural part of conversation.
So it is no surprise that many people have difficulty switching gears to literal mode when interacting with these models.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:16:52.000Z","created_at_i":1774981012,"id":47591378,"options":[],"parent_id":47586380,"points":null,"story_id":47584540,"text":"AI is better at writing prompts than most humans. It requires work and lots of developers don\u2019t think getting good at prompting actually matters.
At least half of the complaints I see on HN boil down to the person's prompts suck. Or the expectation that AI can read their mind.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:30:39.000Z","created_at_i":1774960239,"id":47586380,"options":[],"parent_id":47585940,"points":null,"story_id":47584540,"text":"All these companies use AIs for writing these prompts.
But AI aren't actually very good at writing prompts imo. Like they are superficially good in that they seem to produce lots of vaguely accurate and specific text. And you would hope the specificity would mean it's good.
But they sort of don't capture intent very well. Nor do they seem to understand the failure modes of AI. The "-- describe only what the code change does" is a good example. This is specifc but it also distinctly seems like someone who doesn't actually understand what makes AI writing obvious.
If you compare that vs human written prose about what makes AI writing feel AI you would see the difference. https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing
The above actually feels like text from someone who has read and understands what makes AI writing AI.","title":null,"type":"comment","url":null},{"author":"fleebee","children":[],"created_at":"2026-03-31T16:04:56.000Z","created_at_i":1774973096,"id":47589443,"options":[],"parent_id":47585940,"points":null,"story_id":47584540,"text":"That's not supposed to be surprising. They're dogfooding CC to develop CC. I assume any and every line in this repo is AI generated.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:46:13.000Z","created_at_i":1774957573,"id":47585940,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"> Write commit messages as a human developer would \u2014 describe only what the code change does.
The undercover mode prompt was generated using AI.","title":null,"type":"comment","url":null},{"author":"sandos","children":[{"author":"sgc","children":[{"author":"cdelsolar","children":[],"created_at":"2026-03-31T15:02:05.000Z","created_at_i":1774969325,"id":47588362,"options":[],"parent_id":47586568,"points":null,"story_id":47584540,"text":"lol?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:46:01.000Z","created_at_i":1774961161,"id":47586568,"options":[],"parent_id":47585988,"points":null,"story_id":47584540,"text":"I had a problem to fix and one not only mentioned these "logs", but went on about things like "config", "tests", and a bunch of other unimportant nonsense words. It even went on to point me towards the "manual". Totally robotic monstrosity.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:49:56.000Z","created_at_i":1774957796,"id":47585988,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"This is my pet peeve with LLMs, they almost always fails to write like a normal human would. Mentioning logs, or other meta-things which is not at all interesting.","title":null,"type":"comment","url":null},{"author":"lazysheepherd","children":[{"author":"BoppreH","children":[],"created_at":"2026-03-31T15:02:28.000Z","created_at_i":1774969348,"id":47588376,"options":[],"parent_id":47587505,"points":null,"story_id":47584540,"text":"1) It's not clear to me that this is only for internal tooling, as opposed to publishing commits on public GitHub repos. 2) Yes, it does explicitly say to pretend to be a human. From the link on my post:
> NEVER include in commit messages or PR descriptions:
> [...]
> - The phrase "Claude Code" or any mention that you are an AI","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:58:32.000Z","created_at_i":1774965512,"id":47587505,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"1) This seems to be for strictly Antrophic interal tooling\n2) It does not "pretend to be human" it is instructed to "Write commit messages as a human developer would \u2014 describe only what the code change does."
Since when "describe only what the code change does" is pretending to be human?
You guys are just mining for things to moan about at this point.","title":null,"type":"comment","url":null},{"author":"LelouBil","children":[],"created_at":"2026-03-31T13:59:00.000Z","created_at_i":1774965540,"id":47587517,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"Time to ask if the contributor know what a Capybara is as a new Turing test","title":null,"type":"comment","url":null},{"author":"skeledrew","children":[],"created_at":"2026-03-31T17:11:34.000Z","created_at_i":1774977094,"id":47590464,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"Heh, this is what people who are hostile against AI-generated contributions get. I always figured it'd happen soon enough, and here it is in the wild. Who knows where else it's happening...","title":null,"type":"comment","url":null},{"author":"jen20","children":[{"author":"embedding-shape","children":[],"created_at":"2026-03-31T18:02:00.000Z","created_at_i":1774980120,"id":47591192,"options":[],"parent_id":47590751,"points":null,"story_id":47584540,"text":"Yeah, that was my reaction too. A shame they try to hide themselves, but even worse, the instructions to this "Fake Human" is wrong too!","title":null,"type":"comment","url":null},{"author":"nightpool","children":[{"author":"joquarky","children":[{"author":"nightpool","children":[{"author":"joquarky","children":[],"created_at":"2026-04-06T19:13:13.000Z","created_at_i":1775502793,"id":47665469,"options":[],"parent_id":47594288,"points":null,"story_id":47584540,"text":"Try meditating until you discover the source of internal dialogue.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:25:15.000Z","created_at_i":1774995915,"id":47594288,"options":[],"parent_id":47592858,"points":null,"story_id":47584540,"text":"No, I think a lot of humans can explain why they're adding a new button to the checkout page, or why they're removing a line from the revenue reconciliation job. There's always a reason a change gets made, or else nobody would be working on it at all :)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:15:31.000Z","created_at_i":1774988131,"id":47592858,"options":[],"parent_id":47591652,"points":null,"story_id":47584540,"text":"The question "why" is always answered with post-hoc rationalizations. This applies to both LLMs and humans.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:40:24.000Z","created_at_i":1774982424,"id":47591652,"options":[],"parent_id":47590751,"points":null,"story_id":47584540,"text":"Yes, this is a trend I've noticed strongly with Claude code\u2014it really struggles to explain why. Especially in PR descriptions, it has a strong bias to just summarize the commits and not explain at all why the PR exists.","title":null,"type":"comment","url":null},{"author":"hombre_fatal","children":[],"created_at":"2026-03-31T19:22:55.000Z","created_at_i":1774984975,"id":47592182,"options":[],"parent_id":47590751,"points":null,"story_id":47584540,"text":"I wager that "describe only what the code change does" was someone's attempt to invert "don't add the extra crap you often try to write", not some 4d chess instruction that makes claude larp like a human writing a crappy commit message.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:30:27.000Z","created_at_i":1774978227,"id":47590751,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"Also unintentionally reveals something:
> Write commit messages as a human developer would \u2014 describe only what the code change does.
That's not what a commit message is for, that's what the diff is for. The commit message should explain WHY.
Sadly not doing that likely does indeed make it appear more human...","title":null,"type":"comment","url":null},{"author":"erisnet","children":[],"created_at":"2026-03-31T17:58:34.000Z","created_at_i":1774979914,"id":47591158,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"The first two zips I download today were 9.887.340 bytes,\nwhy is yours 10.222.630 bytes?","title":null,"type":"comment","url":null},{"author":"morissette","children":[],"created_at":"2026-03-31T19:25:37.000Z","created_at_i":1774985137,"id":47592211,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"I pretend to be human most days. I call it the daily facade of who I want to be on a given day. Oh humanity.","title":null,"type":"comment","url":null},{"author":"jacquesm","children":[],"created_at":"2026-03-31T19:40:21.000Z","created_at_i":1774986021,"id":47592393,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"I am Jacques' complete lack of surprise.","title":null,"type":"comment","url":null},{"author":"neilv","children":[],"created_at":"2026-03-31T19:42:10.000Z","created_at_i":1774986130,"id":47592410,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"That's gonna need an explanation. From the ethics/safety/alignment people.","title":null,"type":"comment","url":null},{"author":"dang","children":[],"created_at":"2026-03-31T20:58:54.000Z","created_at_i":1774990734,"id":47593396,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"(We detached this subthread from https://news.ycombinator.com/item?id=47584683.)","title":null,"type":"comment","url":null},{"author":"anabis","children":[],"created_at":"2026-04-01T06:31:06.000Z","created_at_i":1775025066,"id":47597575,"options":[],"parent_id":47585596,"points":null,"story_id":47584540,"text":"Yeah, I thought Anthropic was for AI safety. Telling Ai to not be honest is a bad sign.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:06:39.000Z","created_at_i":1774955199,"id":47585596,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Undercover mode also pretends to be human, which I'm less ok with:
https://github.com/chatgptprojects/claude-code/blob/642c7f94...","title":null,"type":"comment","url":null},{"author":"kschiffer","children":[{"author":"spoiler","children":[{"author":"moron4hire","children":[{"author":"bonoboTP","children":[{"author":"Gigachad","children":[{"author":"andrekandre","children":[{"author":"Gigachad","children":[],"created_at":"2026-04-01T03:25:01.000Z","created_at_i":1775013901,"id":47596431,"options":[],"parent_id":47594917,"points":null,"story_id":47584540,"text":"Agreed. Even in low value stuff I\u2019d so much rather use basic stock images, ms paint drawings or almost anything over AI images. Seeing them is almost like being near someone who stinks or is sick/coughing. It\u2019s a very visceral reaction.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:39:37.000Z","created_at_i":1775000377,"id":47594917,"options":[],"parent_id":47586094,"points":null,"story_id":47584540,"text":"
> I think less of someone as a person if they send me AI slop.\nthere is the fact that such images 'look ai' but even if it were perfect, idk somehow i feel cheap doing that.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:01:48.000Z","created_at_i":1774958508,"id":47586094,"options":[],"parent_id":47586056,"points":null,"story_id":47584540,"text":"Not really. The scale is entirely different. I think less of someone as a person if they send me AI slop.","title":null,"type":"comment","url":null},{"author":"trial3","children":[{"author":"runarberg","children":[],"created_at":"2026-03-31T19:14:09.000Z","created_at_i":1774984449,"id":47592085,"options":[],"parent_id":47586548,"points":null,"story_id":47584540,"text":"An excellent observation. When films became digital the real backlash came when they stopped distributing film for the old film projectors and every movie theaters had to invest in a very expensive DCP projectors. Some couldn\u2019t and were forced to shut down.
If I had lost my local movie theater because of digital film, I would have a really good reason to hate the technology, even though the blame is on the studios forcing that technology on everyone.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:44:42.000Z","created_at_i":1774961082,"id":47586548,"options":[],"parent_id":47586056,"points":null,"story_id":47584540,"text":"perhaps one important detail is that cassette tape guys and Lucasfilm aren\u2019t/weren\u2019t demanding a complete and total restructuring of the economy and society","title":null,"type":"comment","url":null},{"author":"gunsle","children":[{"author":"sanex","children":[],"created_at":"2026-03-31T13:15:04.000Z","created_at_i":1774962904,"id":47586907,"options":[],"parent_id":47586611,"points":null,"story_id":47584540,"text":"Project Hail Mary is a great example of not relying on CGI.","title":null,"type":"comment","url":null},{"author":"xnorswap","children":[{"author":"andrekandre","children":[],"created_at":"2026-03-31T23:44:57.000Z","created_at_i":1775000697,"id":47594956,"options":[],"parent_id":47586991,"points":null,"story_id":47584540,"text":"it may be an unpopular opinion but i feel like that watching any of the marvel movies... its like its just a showcase for green screens and ridiculous rubber-band acrobatics cgi everywhere...
that kind if stuff might work in anime or cartoons, but live action just looks ridiculous to me for the most part.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:22:17.000Z","created_at_i":1774963337,"id":47586991,"options":[],"parent_id":47586611,"points":null,"story_id":47584540,"text":"Not just in the obvious ways either, even good CGI has been detrimental to the film (and TV) making process.
I was watching some behind the scenes footage from something recently, and the thing that struck me most was just how they wouldn't bother with the location shoot now and just green-screen it all for the convenience.
Even good CGI is changing not just how films are made, but what kinds of films get shot and what kind of stories get told.
Regardless of the quality of the output, there's a creativeness in film-making that is lost as CGI gets better and cheaper to do.","title":null,"type":"comment","url":null},{"author":"Levitz","children":[],"created_at":"2026-03-31T13:43:12.000Z","created_at_i":1774964592,"id":47587291,"options":[],"parent_id":47586611,"points":null,"story_id":47584540,"text":""Literally everyone" can't even agree on whether Polio is bad.
I myself would disagree that CGI itself is a bad thing.","title":null,"type":"comment","url":null},{"author":"NitpickLawyer","children":[],"created_at":"2026-03-31T13:47:58.000Z","created_at_i":1774964878,"id":47587359,"options":[],"parent_id":47586611,"points":null,"story_id":47584540,"text":"Anecdata-- from me. I think cgi can be a net positive.","title":null,"type":"comment","url":null},{"author":"delecti","children":[],"created_at":"2026-03-31T14:22:49.000Z","created_at_i":1774966969,"id":47587824,"options":[],"parent_id":47586611,"points":null,"story_id":47584540,"text":"I could maybe agree in the sense of "has had detrimental effects", but certainly not in the sense of "net detrimental".","title":null,"type":"comment","url":null},{"author":"CamperBob2","children":[{"author":"runarberg","children":[{"author":"CamperBob2","children":[{"author":"jamiek88","children":[{"author":"CamperBob2","children":[{"author":"andrekandre","children":[{"author":"runarberg","children":[],"created_at":"2026-03-31T23:58:54.000Z","created_at_i":1775001534,"id":47595069,"options":[],"parent_id":47595005,"points":null,"story_id":47584540,"text":"My guess is that post-training has gotten a lot better in the last couple of years and what people are attributing to better models are actually just traditional (non-LLM) models they place on top of the LLM which makes it appears that the model has increased in quality (including by seemingly fewer hallucination).
If this is the case it would be observed with different prompting strategies, when you find a prompt which puts more weight on the post-training models.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:50:47.000Z","created_at_i":1775001047,"id":47595005,"options":[],"parent_id":47594633,"points":null,"story_id":47584540,"text":"i think they are referring to statements that they have "solved" hallucinations and it wont be a problem anymore (which it obviously isn't yet anyways)
[1] https://news.ycombinator.com/item?id=44779198","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:03:27.000Z","created_at_i":1774998207,"id":47594633,"options":[],"parent_id":47594493,"points":null,"story_id":47584540,"text":"Hallucinations have been solved?!
Apparently not, because no one but you implied that they had been.
There are prompting strategies that improve the odds greatly, but like the GGP, you've made up your mind, so it's a waste of time to argue otherwise.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:48:20.000Z","created_at_i":1774997300,"id":47594493,"options":[],"parent_id":47594204,"points":null,"story_id":47584540,"text":"Hallucinations have been solved?! That\u2019s great news! Must have missed that.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:16:02.000Z","created_at_i":1774995362,"id":47594204,"options":[],"parent_id":47592126,"points":null,"story_id":47584540,"text":"But you can\u2019t prompt your way out of hallucinations and other AI artifacts
That's not the case, and hasn't been for some time, but it sounds like your mind's made up.","title":null,"type":"comment","url":null},{"author":"Ferret7446","children":[{"author":"runarberg","children":[{"author":"moron4hire","children":[],"created_at":"2026-04-01T10:43:23.000Z","created_at_i":1775040203,"id":47599131,"options":[],"parent_id":47596534,"points":null,"story_id":47584540,"text":"I had a very similar experience, looking for music to play during D&D sessions. Not paying close attention to the music, it seemed like it fit the bill. Once I started listening more closely, there were lots of issues that became readily apparent.
My dad has also started sharing with me links on Facebook to pop songs that have been re-arranged in different genres. This was a big area of fun for a number of folks in my family several years ago as we discovered YouTube artists like Chase Holfelder who put significant effort into making very high quality rearrangements. But I kept noticing these weird issues in the new songs.
I've gotten to where I can identify an AI generated song almost immediately: there's a weird, high frequency hiss in the mix that sounds like heavy noise getting to overcome compression artifacts but the source from which it's coming should be clean. There's a general lack of enthusiasm to the lyrics and a boring, nonsensical progression to the lyrics on original arrangements. Sometimes, the person generating the song tries to hide that last issue by generating instrumentals only or they use one of those try-to-hard-to-sound-badass Country Rock genres that are popular on Tik Tok to stick on top of clips from the TV show Yellowstone (WTF is with that?!), but then when I check the details, there's an obviously AI cover art for artists I've never heard of. The accounts will be anthologies full of these artists that have never existed.
So, I know people keep parroting "a good artist can use any tool". But I've yet to see it. All this "democratizing art" (didn't know anyone was gate keeping it to begin with, certainly have not seen any lack of talent online in several years) doesn't seem to be producing results. It becomes pretty obvious very quickly it's all just a pump and dump scheme to Get Them Clicks.","title":null,"type":"comment","url":null},{"author":"Ferret7446","children":[{"author":"moron4hire","children":[{"author":"runarberg","children":[],"created_at":"2026-04-03T01:49:52.000Z","created_at_i":1775180992,"id":47622417,"options":[],"parent_id":47616964,"points":null,"story_id":47584540,"text":"I think it is worse then a No True Scotsman. I think your parent actually performed a category mistake here. Survivorship bias does not apply here. Whether or not I notice or even unknowingly enjoy AI generated content is not in the same category as how much I notice or enjoy CGI.
The difference is in the authorship. Actual work and skill goes into CGI, and people generally notice bad CGI, and it generally affects how you judge the art. Sometimes CGI is actually part of the art and you are supposed to notice it, and it is still good (think how Cher use Autotune in Do You Believe). There is no such equivalence with AI.
To further elaborate. Bad CGI is often (but not always) used as a cost-cutting means. Directors (or producers encourage directors to) use it when they want to save money on practical effects or even cover up mistakes that happened during shooting and want to avoid an expensive re-shoot. This can work OK if used sparingly and carefully, however if this is done a lot and without the needed care, you will notice it, and you will judge the work from it. AI content is kind of like that, except that is kind of all what AI is. The other couldn\u2019t be bothered to do the work and just prompted an AI to do it for them.
To summarize: AI is not like CGI in general, it is much closer to a strict subset of CGI which only includes bad CGI.","title":null,"type":"comment","url":null}],"created_at":"2026-04-02T16:54:39.000Z","created_at_i":1775148879,"id":47616964,"options":[],"parent_id":47608472,"points":null,"story_id":47584540,"text":"This sounds dangerously close to a No True Scotsman argument. Any example one could provide, you've teed it up nicely to claim that no, you didn't mean that one, obviously, because you could tell. No, it's some other thing that you haven't found yet. That's the passing-AI.","title":null,"type":"comment","url":null}],"created_at":"2026-04-02T00:21:38.000Z","created_at_i":1775089298,"id":47608472,"options":[],"parent_id":47596534,"points":null,"story_id":47584540,"text":"You don't understand. I mean content that even now, you don't know it is AI.
Obviously you think the AI content that you can identify is bad. But there is content you've encountered that you think is good and not AI content, that actually is AI generated.
That's the survivorship bias.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T03:39:58.000Z","created_at_i":1775014798,"id":47596534,"options":[],"parent_id":47596000,"points":null,"story_id":47584540,"text":"I have and I hated it.
The story is that I was getting into a new genre of music, namely Japanese City pop from the 1980s. I was totally unfamiliar with the genre and started listening to it on YouTube. I found one playlist, which I listened to a lot, thinking: \u201cwow, this is very formulaic, and the lyrics are very generic\u201d but I kind of thought that was just how the genre went. Finally had planned to use it for during a small local event, but when I went to find out who the artists were I embarrassingly found out it was all AI generated.
Thing is, in this instance I knew nothing of the source material, when I went to get actual songs, written by actual people, the difference was start. I would be able to recognize AI generated City pop in an instant now 8 months later. This experience kind of felt like I had been scammed. That my ignorance of the genre had been taken advantage of. It was not pleasant.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T02:19:15.000Z","created_at_i":1775009955,"id":47596000,"options":[],"parent_id":47592126,"points":null,"story_id":47584540,"text":"I guarantee you have encountered AI content and not realized it was AI. I assume you've heard of the survivorship bias?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:17:54.000Z","created_at_i":1774984674,"id":47592126,"options":[],"parent_id":47589850,"points":null,"story_id":47584540,"text":"Not the same. The more effort you put into CGI the more invisible it becomes. But you can\u2019t prompt your way out of hallucinations and other AI artifacts. AI is a completely different technology from CGI. There is no equivalence between them.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:29:16.000Z","created_at_i":1774974556,"id":47589850,"options":[],"parent_id":47586611,"points":null,"story_id":47584540,"text":"90% of the time, you wouldn't know CGI if you saw it. That's the 'good' CGI.
Same thing is true of AI output.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:49:59.000Z","created_at_i":1774961399,"id":47586611,"options":[],"parent_id":47586056,"points":null,"story_id":47584540,"text":"I think literally everyone could agree CGI has been detrimental to the quality of films.","title":null,"type":"comment","url":null},{"author":"runarberg","children":[],"created_at":"2026-03-31T19:10:09.000Z","created_at_i":1774984209,"id":47592033,"options":[],"parent_id":47586056,"points":null,"story_id":47584540,"text":"It is not. People resisted bad CGI. During the advent of CGI people celebrated the masterpiece of the Matrix and even Titanic. They hated however the Scorpion King.","title":null,"type":"comment","url":null},{"author":"moron4hire","children":[],"created_at":"2026-04-01T00:06:29.000Z","created_at_i":1775001989,"id":47595125,"options":[],"parent_id":47586056,"points":null,"story_id":47584540,"text":"No, I don't think most people are really against AI Gen works "on principle". Or at least not in any interpretation of "on principle" that would allow for you to be dismissive of complaints in this way.
I think principles are important. Especially when it comes to art, principle might be all we have. Going back to the crypto example, NFTs were art that real people had made. In some cases, very good art. People railed against NFTs despite the quality of the art. That is being against something on-principle. Comparatively, if my local grocery chains were owned by neonazis, I'd have a much harder time of standing on principle, giving that doing so may have a negative impact on my ability to survive and prosper.
AI Gen works, on the other hand, most often do not come with readily available marking that it is AI Gen. What people are complaining about is the lack of quality in the work. If they accuse a poorly human-written article of being AI Gen, that's just a mistake. But the general case is a legitimate evaluation of the quality of the material and the conditions under which it was made and presented.
In my own case, while I certainly have plenty of "principled" reasons to dislike AI Gen works, I also dislike it because it's just garbage. Oh yeah, sure, it's impressive that a computer can spit out reasonable content at all. It would equally be impressive for a chimpanzee to start talking in full sentences. That doesn't mean I'm going to start going to the chimpanzee for dissertations on the human condition.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:56:36.000Z","created_at_i":1774958196,"id":47586056,"options":[],"parent_id":47585876,"points":null,"story_id":47584540,"text":"It's how people resisted CGI back in the day. What people dislike is low quality. There is a loud subset who are really against it on principle like we also have people who insist on analog music but regular people are much more practical but they don't post about this all day on the internet.","title":null,"type":"comment","url":null},{"author":"sunaookami","children":[{"author":"neutronicus","children":[],"created_at":"2026-03-31T12:47:34.000Z","created_at_i":1774961254,"id":47586584,"options":[],"parent_id":47586067,"points":null,"story_id":47584540,"text":"Especially true in gaming communities.
IMO it's a combination of long-running paranoia about cost-cutting and quality, and a sort of performative allegiance to artists working in the industry.","title":null,"type":"comment","url":null},{"author":"Ferret7446","children":[],"created_at":"2026-04-01T02:21:40.000Z","created_at_i":1775010100,"id":47596019,"options":[],"parent_id":47586067,"points":null,"story_id":47584540,"text":"And E33 is also a good example that these users are a minority and effectively immaterial. They don't affect sales or the popular opinion.
People don't care about AI. They only care whether the product is good.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:57:52.000Z","created_at_i":1774958272,"id":47586067,"options":[],"parent_id":47585876,"points":null,"story_id":47584540,"text":"No there is a very loud minority of users who are very anti AI that hate on anything that is even remotely connected to AI and let everyone know with false claims. See the game Expedition 33 for example.","title":null,"type":"comment","url":null},{"author":"Levitz","children":[],"created_at":"2026-03-31T13:46:49.000Z","created_at_i":1774964809,"id":47587345,"options":[],"parent_id":47585876,"points":null,"story_id":47584540,"text":"And yet, no game has problems selling due to these reactions. As a matter of fact, the vast majority of people can't even tell if AI has been used here or there unless told.
I reckon it's just drama paraded by gaming "journalists" and not much else. You will find people expressing concern on Reddit or Bluesky, but ultimately it doesn't matter.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:40:10.000Z","created_at_i":1774957210,"id":47585876,"options":[],"parent_id":47585787,"points":null,"story_id":47584540,"text":"To me, this is a sign of just how much regular people do not want AI. This is worse than crypto and metaverse before it. Crypto, people could ignore and the dumb ape pictures helped you figure out who to avoid. Metaverse, some folks even still enjoyed VR and AR without the digital real estate bullshit. And neither got shoved down your throat in everyday, mundane things like writing a paper in Word or trying to deal with your auto mechanic.
But AI is causing such visceral reactions that it's bleeding into other areas. People are so averse to AI they don't mind a few false positives.","title":null,"type":"comment","url":null},{"author":"PunchyHamster","children":[{"author":"foobarchu","children":[],"created_at":"2026-04-01T00:00:04.000Z","created_at_i":1775001604,"id":47595075,"options":[],"parent_id":47587500,"points":null,"story_id":47584540,"text":"> all that needs to be done
The honor system is never a sustainable solution. It's not even down to corporate greed, it's just not something that works at scale, especially when there's money to be made, and even more especially when there isn't.","title":null,"type":"comment","url":null},{"author":"Ferret7446","children":[],"created_at":"2026-04-01T02:15:16.000Z","created_at_i":1775009716,"id":47595977,"options":[],"parent_id":47587500,"points":null,"story_id":47584540,"text":"And all that's needed for world peace is for everyone to stop fighting.
Which of course won't happen because we live in reality and not fantasy where we can dream that "people should just do X"","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:58:12.000Z","created_at_i":1774965492,"id":47587500,"options":[],"parent_id":47585787,"points":null,"story_id":47584540,"text":"All that is needed to solve that is to reliably put AI disclaimer on things done by AI
Which of course won't be done because corporations don't want that (except Valve I guess), so blame them.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:27:17.000Z","created_at_i":1774956437,"id":47585787,"options":[],"parent_id":47585627,"points":null,"story_id":47584540,"text":"Random aside: I've seen a 2015 game be accused of AI slop on Steam because it used a similar concept... And mind you, there's probably thousands of games that do this.
First it was punctuation and grammar, then linguistic coherence, and now it's tiny bits of whimsy that are falling victim to AI accusations. Good fucking grief","title":null,"type":"comment","url":null},{"author":"bonoboTP","children":[],"created_at":"2026-03-31T11:53:18.000Z","created_at_i":1774957998,"id":47586020,"options":[],"parent_id":47585627,"points":null,"story_id":47584540,"text":"It's not hard to find them, they are in clear text in the binary, you can search for known ones with grep and find the rest nearby. You could even replace them inplace (but now its configurable).","title":null,"type":"comment","url":null},{"author":"Gormo","children":[{"author":"avaer","children":[{"author":"ticulatedspline","children":[],"created_at":"2026-03-31T12:26:25.000Z","created_at_i":1774959985,"id":47586325,"options":[],"parent_id":47586216,"points":null,"story_id":47584540,"text":"Def not alone","title":null,"type":"comment","url":null},{"author":"NooneAtAll3","children":[{"author":"lanyard-textile","children":[],"created_at":"2026-04-01T07:32:32.000Z","created_at_i":1775028752,"id":47597960,"options":[],"parent_id":47596092,"points":null,"story_id":47584540,"text":""Reticulating Splines" was one of many possible loading screen phrases for games like The Sims and Sim City.
Instead of realistic ones like "Loading assets" and "Reading file", they would give humorous nonsensical ones like "Reticulating Splines".","title":null,"type":"comment","url":null},{"author":"wernsey","children":[],"created_at":"2026-04-01T09:50:45.000Z","created_at_i":1775037045,"id":47598810,"options":[],"parent_id":47596092,"points":null,"story_id":47584540,"text":"SimCity 2000 showed a message "Reticulating splines" when it generated the landscape that didn't mean anything in particular.
It became a bit of a meme in the years since.","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T02:34:20.000Z","created_at_i":1775010860,"id":47596092,"options":[],"parent_id":47586216,"points":null,"story_id":47584540,"text":"is it a reference to something?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:15:57.000Z","created_at_i":1774959357,"id":47586216,"options":[],"parent_id":47586162,"points":null,"story_id":47584540,"text":"Relieved to know I'm not the only one who grepped for that. Thank you for making me feel sane, friend.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:09:58.000Z","created_at_i":1774958998,"id":47586162,"options":[],"parent_id":47585627,"points":null,"story_id":47584540,"text":"I'm glad "reticulating" is in there. Just need to make sure "splines" is in the nouns list!","title":null,"type":"comment","url":null},{"author":"moontear","children":[{"author":"tommit","children":[],"created_at":"2026-03-31T12:25:43.000Z","created_at_i":1774959943,"id":47586313,"options":[],"parent_id":47586181,"points":null,"story_id":47584540,"text":"oh wow, there are like 10 opened every minute. seems spam-y","title":null,"type":"comment","url":null},{"author":"avaer","children":[{"author":"lanbin","children":[],"created_at":"2026-03-31T16:12:41.000Z","created_at_i":1774973561,"id":47589569,"options":[],"parent_id":47586402,"points":null,"story_id":47584540,"text":"\u5408\u5f71 doesn't sound like any Chinese slang. That is just what "group photo" means.","title":null,"type":"comment","url":null},{"author":"breakds","children":[],"created_at":"2026-03-31T18:29:39.000Z","created_at_i":1774981779,"id":47591519,"options":[],"parent_id":47586402,"points":null,"story_id":47584540,"text":"In this situation, it means "Hey I have been here and observed this!"","title":null,"type":"comment","url":null},{"author":"Ferret7446","children":[],"created_at":"2026-04-01T02:12:02.000Z","created_at_i":1775009522,"id":47595950,"options":[],"parent_id":47586402,"points":null,"story_id":47584540,"text":"GitHub had always been about social coding from the beginning (e.g., forking and PRing repos)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:32:36.000Z","created_at_i":1774960356,"id":47586402,"options":[],"parent_id":47586181,"points":null,"story_id":47584540,"text":"It seems human. It taught me \u5408\u5f71, which seems to be Chinese slang for just wanting to be in the comments. Probably not a coincidence that it's after work time in China.
Really interesting to see Github turn into 4chan for a minute, like GH anons rolling for trips.","title":null,"type":"comment","url":null},{"author":"Quarrel","children":[],"created_at":"2026-03-31T12:36:20.000Z","created_at_i":1774960580,"id":47586447,"options":[],"parent_id":47586181,"points":null,"story_id":47584540,"text":"trying to get github to nuke the repo? at a guess.
certainly nothing friendly.","title":null,"type":"comment","url":null},{"author":"proactivesvcs","children":[],"created_at":"2026-03-31T12:43:43.000Z","created_at_i":1774961023,"id":47586527,"options":[],"parent_id":47586181,"points":null,"story_id":47584540,"text":"I saw this on restic's main repository the other day.","title":null,"type":"comment","url":null},{"author":"g947o","children":[],"created_at":"2026-03-31T12:43:44.000Z","created_at_i":1774961024,"id":47586531,"options":[],"parent_id":47586181,"points":null,"story_id":47584540,"text":"There have been massive GitHub issue spams recently, including in Microsoft's WSL repository.
https://github.com/microsoft/WSL/issues/40028","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:11:58.000Z","created_at_i":1774959118,"id":47586181,"options":[],"parent_id":47585627,"points":null,"story_id":47584540,"text":"What's going on with the issues in that repo? https://github.com/instructkr/claude-code/issues","title":null,"type":"comment","url":null},{"author":"world2vec","children":[{"author":"raesene9","children":[],"created_at":"2026-03-31T13:14:12.000Z","created_at_i":1774962852,"id":47586893,"options":[],"parent_id":47586543,"points":null,"story_id":47584540,"text":"I think the original repo OP mentioned decided not to host the code any more, but given there are 28k+ forks, it's not too hard to find again...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:44:27.000Z","created_at_i":1774961067,"id":47586543,"options":[],"parent_id":47585627,"points":null,"story_id":47584540,"text":"Did they remove that in some very recent commit?","title":null,"type":"comment","url":null},{"author":"tony-vlcek","children":[{"author":"bigwheels","children":[],"created_at":"2026-03-31T20:38:18.000Z","created_at_i":1774989498,"id":47593163,"options":[],"parent_id":47589180,"points":null,"story_id":47584540,"text":"Also discoverable via:
strings $(which claude) | grep 'Swirling'Here's one that works (for now):\nhttps://github.com/chatgptprojects/claude-code/blob/642c7f94...","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:10:49.000Z","created_at_i":1774955449,"id":47585627,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Finally all spinner verbs revealed:\nhttps://github.com/instructkr/claude-code/blob/main/src/cons...","title":null,"type":"comment","url":null},{"author":"tekacs","children":[{"author":"panny","children":[],"created_at":"2026-03-31T11:45:12.000Z","created_at_i":1774957512,"id":47585930,"options":[],"parent_id":47585763,"points":null,"story_id":47584540,"text":"Too little to late. Someone has it building now.
https://github.com/oboard/claude-code-rev","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:24:41.000Z","created_at_i":1774956281,"id":47585763,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"In the app, it now reads:
> current: 2.1.88 \u00b7 latest: 2.1.87
Which makes me think they pulled it - although it still shows up as 2.1.88 on npmjs for now (cached?).","title":null,"type":"comment","url":null},{"author":"hk__2","children":[{"author":"ankaz","children":[{"author":"solaire_oa","children":[{"author":"ankaz","children":[],"created_at":"2026-04-01T00:05:13.000Z","created_at_i":1775001913,"id":47595113,"options":[],"parent_id":47589776,"points":null,"story_id":47584540,"text":"The source repo doesn't have a package.json, so I extracted the version directly from the binary (~/.local/share/claude/versions/2.1.87)
Axios sets a VERSION constant that it uses in user-agent headers, boundaries and errors. I scanned the binary for all references like axios, isAxiosError and AxiosError - the code references the same variable namespace (X1H, Tj, eq), suggesting a single bundled copy. In the minified bundle, that VERSION constant was stored in a variable called X1H. Searching the binary for all references to X1H confirms it's only used in axios contexts:
var X1H="1.13.6"\n E.set("User-Agent","axios/"+X1H, ...)\n {tag:`axios-${X1H}-boundary`, ...}\n "[Axios v"+X1H+"] Transitional option ..."\n Tj.VERSION=X1H; Tj.AxiosError=eq; Tj.CancelToken=...\nTo stop Claude Code from auto-updating, add `export DISABLE_AUTOUPDATER=1` to your global environment variables (~/.bashrc, ~/.zshrc, or such), restart all sessions and check that it works with `claude doctor`, it should show `Auto-updates: disabled (DISABLE_AUTOUPDATER set)`","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:31:06.000Z","created_at_i":1774956666,"id":47585810,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"For a combo with another HN homepage story, Claude Code uses\u2026 Axios: https://x.com/icanvardar/status/2038917942314778889?s=20
https://news.ycombinator.com/item?id=47582220","title":null,"type":"comment","url":null},{"author":"phtrivier","children":[{"author":"treexs","children":[],"created_at":"2026-03-31T11:48:06.000Z","created_at_i":1774957686,"id":47585962,"options":[],"parent_id":47585855,"points":null,"story_id":47584540,"text":"Yes it's the claude code CLI tool / coding agent harness, not the weights.
This code hasn't been open source until now and contains information like the system prompts, internal feature flags, etc.","title":null,"type":"comment","url":null},{"author":"59nadir","children":[{"author":"phtrivier","children":[{"author":"59nadir","children":[],"created_at":"2026-03-31T16:30:49.000Z","created_at_i":1774974649,"id":47589874,"options":[],"parent_id":47588502,"points":null,"story_id":47584540,"text":"No, I meant that it is other people's code run through a tumbler, as is all LLM-generated code.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:10:02.000Z","created_at_i":1774969802,"id":47588502,"options":[],"parent_id":47586034,"points":null,"story_id":47584540,"text":"You believe it's just a fake ? (That would be ironic if the fake was generated by... claude itself. Anyway.)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:54:34.000Z","created_at_i":1774958074,"id":47586034,"options":[],"parent_id":47585855,"points":null,"story_id":47584540,"text":"> I don't like reading leaked code
Don't worry about that, the code in that repository isn't Anthropic's to begin with.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:36:54.000Z","created_at_i":1774957014,"id":47585855,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Maybe the OP could clarify, I don't like reading leaked code, but I'm curious: \nmy understanding is that is it the source code for "claude code", the coding assistant that remotely calls the LLMs.
Is that correct ? The weights of the LLMs are _not_ in this repo, right ?
It sure sucks for anthropic to get pawned like this, but it should not affect their bottom line much ?","title":null,"type":"comment","url":null},{"author":"Painsawman123","children":[{"author":"MallocVoidstar","children":[],"created_at":"2026-03-31T11:56:15.000Z","created_at_i":1774958175,"id":47586052,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"Codex is open source: https://github.com/openai/codex","title":null,"type":"comment","url":null},{"author":"kaszanka","children":[],"created_at":"2026-03-31T11:59:56.000Z","created_at_i":1774958396,"id":47586082,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"Is https://github.com/google-gemini/gemini-cli not 'the flagship agent' itself? It looks that way to me, for example here's a part of the prompt https://github.com/google-gemini/gemini-cli/blob/e293424bb49...","title":null,"type":"comment","url":null},{"author":"loveparade","children":[],"created_at":"2026-03-31T12:03:19.000Z","created_at_i":1774958599,"id":47586106,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"It's exactly the same as the open source codex/gemini and other clis like opencode. There is no secret sauce in the claude cli, and the agent harness itself is no better (worse IMO) than the others. The only thing interesting about this leak is that it may contain unreleased features/flags that are not public yet and hint at what Anthropic is working on.","title":null,"type":"comment","url":null},{"author":"hmokiguess","children":[],"created_at":"2026-03-31T13:57:07.000Z","created_at_i":1774965427,"id":47587480,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"Do you think the other companies don\u2019t have sufficient resources to attempt reverse engineering and deobfuscating a client side application?
The source maps help for sure, but it\u2019s not like client code is kept secret, maybe they even knew about the source maps a while back just didn\u2019t bother making it common knowledge.
This is not a leak of the model weights or server side code.","title":null,"type":"comment","url":null},{"author":"mmaunder","children":[],"created_at":"2026-03-31T14:57:34.000Z","created_at_i":1774969054,"id":47588283,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"Agreed. This is a big deal.","title":null,"type":"comment","url":null},{"author":"nunez","children":[],"created_at":"2026-03-31T15:04:29.000Z","created_at_i":1774969469,"id":47588405,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"Yeah, this is the LLaMa leak moment for agentic app dev, IMO. Huge deal. Big win for Opencode and the like.","title":null,"type":"comment","url":null},{"author":"weird-eye-issue","children":[{"author":"ithkuil","children":[{"author":"petcat","children":[],"created_at":"2026-03-31T22:38:18.000Z","created_at_i":1774996698,"id":47594395,"options":[],"parent_id":47589564,"points":null,"story_id":47584540,"text":"> they add more and more features and fix more and more bugs
My experience has been that they add far more bugs in every release than they fix","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:12:25.000Z","created_at_i":1774973545,"id":47589564,"options":[],"parent_id":47588629,"points":null,"story_id":47584540,"text":"yeah it actually works to use claude to reverse engineer itself; I've used that to workaround some problems. E.g. that's how I discovered that I had to put two slashes for absolute paths in sandbox config. The thing is, the claude team is so quick that soon enough they add more and more features and fix more and more bugs that your workarounds become obsolete","title":null,"type":"comment","url":null},{"author":"sodapopcan","children":[{"author":"Capricorn2481","children":[{"author":"sodapopcan","children":[],"created_at":"2026-04-01T17:35:15.000Z","created_at_i":1775064915,"id":47603973,"options":[],"parent_id":47594114,"points":null,"story_id":47584540,"text":"Fair, I was just seeing so much of this and randomly responded to yours.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T22:07:08.000Z","created_at_i":1774994828,"id":47594114,"options":[],"parent_id":47591224,"points":null,"story_id":47584540,"text":"> The point is that a "secure coding platform" leaked something they were trying to keep under wraps, whether the contents of the leak matter or not
Sure, but that's completely different from what they were responding to to, which was someone insinuating the Claude Code CLI has secret sauce that makes it better than the competition.","title":null,"type":"comment","url":null},{"author":"weird-eye-issue","children":[{"author":"sodapopcan","children":[{"author":"weird-eye-issue","children":[],"created_at":"2026-04-02T05:00:30.000Z","created_at_i":1775106030,"id":47610161,"options":[],"parent_id":47604011,"points":null,"story_id":47584540,"text":"As a Claude Code user why should I care?","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T17:38:18.000Z","created_at_i":1775065098,"id":47604011,"options":[],"parent_id":47600658,"points":null,"story_id":47584540,"text":"> No reasonable person actually thinks that Claude Code cannot make mistakes.
This was insanely rookie mistake that could have been caught if anyone was paying attention as opposed to "vibing."","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T13:34:25.000Z","created_at_i":1775050465,"id":47600658,"options":[],"parent_id":47591224,"points":null,"story_id":47584540,"text":"No reasonable person actually thinks that Claude Code cannot make mistakes. So if your point is that this is going to change anybody's opinion about it then I think that's pretty silly.
Also who really cares about the roadmap? Any feature they release can be easily copied quickly. The only moat they have at the moment is in giving access to their models via a subscription.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:04:05.000Z","created_at_i":1774980245,"id":47591224,"options":[],"parent_id":47588629,"points":null,"story_id":47584540,"text":"The point is that a "secure coding platform" leaked something they were trying to keep under wraps, whether the contents of the leak matter or not.
Also, as many others have pointed out, there is roadmap info in here that wouldn't be available in the production build.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:18:31.000Z","created_at_i":1774970311,"id":47588629,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"It doesn't matter that much. Trust me you could just have an LLM reverse engineer the obfuscated code.","title":null,"type":"comment","url":null},{"author":"danmaz74","children":[{"author":"mholm","children":[],"created_at":"2026-03-31T17:23:42.000Z","created_at_i":1774977822,"id":47590648,"options":[],"parent_id":47590173,"points":null,"story_id":47584540,"text":"The only real secret sauce is the training methods and datasets used for refining harness usage. Claude Code is a lot better than gemini-cli/open-code/etc because Claude is specifically trained on how to run in that environment. It's been rlhf'd to use the provided tools correctly, and know the framework in which it operates, instead of relying solely on context.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:51:24.000Z","created_at_i":1774975884,"id":47590173,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"I guess that the most important potential "secret sauce" for a coding agent would be its prompts, but that's also one of the easiest things to find out by simply intercepting its messages.","title":null,"type":"comment","url":null},{"author":"IceWreck","children":[],"created_at":"2026-03-31T18:08:23.000Z","created_at_i":1774980503,"id":47591275,"options":[],"parent_id":47585866,"points":null,"story_id":47584540,"text":"> What Google and OpenAi have open sourced is their Agents SDK, a toolkit, not the secret sauce of how their flagship agents are wired under the hood
And how is that any different? Claude Code is a harness, similar to open source ones like Codex, Gemini CLI, OpenCode etc. Their prompts were already public because you could connect it to your own LLM gateway and see everything. The code was transpiled javascript which is trivial to read with LLMs anyways.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T11:38:33.000Z","created_at_i":1774957113,"id":47585866,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Really surprising how many people are downplaying this leak!\n"Google and OpenAi have already open sourced their Agents, so this leak isn't that relevant " What Google and OpenAi have open sourced is their Agents SDK, a toolkit, not the secret sauce of how their flagship agents are wired under the hood!\nexpect the takedown hammer on the tweet, the R2 link, and any public repos soon","title":null,"type":"comment","url":null},{"author":"Sathwickp","children":[],"created_at":"2026-03-31T11:39:02.000Z","created_at_i":1774957142,"id":47585870,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"They do have a couple of interesting features that has not been publicly heard of yet:
Like KAIROS which seems to be like an inbuilt ai assistant and Ultraplan which seems to enable remote planning workflows, where a separate environment explores a problem, generates a plan, and then pauses for user approval before execution.","title":null,"type":"comment","url":null},{"author":"thefilmore","children":[],"created_at":"2026-03-31T11:49:17.000Z","created_at_i":1774957757,"id":47585979,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"400k lines of code per scc","title":null,"type":"comment","url":null},{"author":"sbochins","children":[],"created_at":"2026-03-31T11:57:01.000Z","created_at_i":1774958221,"id":47586060,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Does this matter? I think every other agent cli is open source. I don\u2019t even know why Anthropic insist upon having theirs be closed source.","title":null,"type":"comment","url":null},{"author":"bdangubic","children":[],"created_at":"2026-03-31T12:05:16.000Z","created_at_i":1774958716,"id":47586125,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I have 705 PRs ready to go :)","title":null,"type":"comment","url":null},{"author":"VadimPR","children":[{"author":"bethekind","children":[],"created_at":"2026-03-31T14:41:13.000Z","created_at_i":1774968073,"id":47588057,"options":[],"parent_id":47586180,"points":null,"story_id":47584540,"text":"This. If I run 4 Claude code opus agents with subagents, my 8gb of RAM just dies.
I know they can do better","title":null,"type":"comment","url":null},{"author":"ex-aws-dude","children":[],"created_at":"2026-03-31T18:32:47.000Z","created_at_i":1774981967,"id":47591553,"options":[],"parent_id":47586180,"points":null,"story_id":47584540,"text":"But its already optimized so well that its comparable to a "small game engine"?
https://nitter.net/trq212/status/2014051501786931427#m","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:11:45.000Z","created_at_i":1774959105,"id":47586180,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Anthropic team does an excellent job of speeding up Claude Code when it slows down, but for the sake of RAM and system resources, it would be nice to see it rewritten in a more performant framework!
And now, with Claude on a Ralph loop, you can.","title":null,"type":"comment","url":null},{"author":"zoobab","children":[],"created_at":"2026-03-31T12:14:26.000Z","created_at_i":1774959266,"id":47586199,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Just a client side written in JS, nothing to see here, the LLM is still secret.
They could have written that in curl+bash that would not have changed much.","title":null,"type":"comment","url":null},{"author":"hemantkamalakar","children":[],"created_at":"2026-03-31T12:39:17.000Z","created_at_i":1774960757,"id":47586473,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Today being March 31st, is this a genuine issue or just perfectly timed April Fools noise? What do you think?","title":null,"type":"comment","url":null},{"author":"hemantkamalakar","children":[],"created_at":"2026-03-31T12:39:28.000Z","created_at_i":1774960768,"id":47586478,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"today being March 31st, is this a genuine issue or just perfectly timed April Fools noise? What do you think?","title":null,"type":"comment","url":null},{"author":"artdigital","children":[],"created_at":"2026-03-31T12:40:47.000Z","created_at_i":1774960847,"id":47586495,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Now waiting for someone to point Codex at it and rebuild a new Claude Code in Golang to see if it would perform better","title":null,"type":"comment","url":null},{"author":"zurfer","children":[{"author":"raesene9","children":[{"author":"tmp10423288442","children":[{"author":"raesene9","children":[],"created_at":"2026-04-03T10:24:00.000Z","created_at_i":1775211840,"id":47624997,"options":[],"parent_id":47594734,"points":null,"story_id":47584540,"text":"but once forked people will have local copies, that can be put up onto other sites, if GH take them down.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:16:10.000Z","created_at_i":1774998970,"id":47594734,"options":[],"parent_id":47586929,"points":null,"story_id":47584540,"text":"Forks are easy for Github to shut down simultaneously. What you really want is to upload the code as a new repo (ideally a different name from the original one). But it shouldn't be too hard in practice to detect uploading the same codebase as one that's taken down if that's desired.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:16:54.000Z","created_at_i":1774963014,"id":47586929,"options":[],"parent_id":47586502,"points":null,"story_id":47584540,"text":"there are a .....lot of forks already, no putting the genie back in the bottle for this one, I'd imagine.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:42:07.000Z","created_at_i":1774960927,"id":47586502,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"too much pressure. the author deleted the real source code: https://github.com/instructkr/claude-code/commit/7c3c5f7eb96...","title":null,"type":"comment","url":null},{"author":"napo","children":[],"created_at":"2026-03-31T12:44:47.000Z","created_at_i":1774961087,"id":47586550,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"The autoDream feature looks interesting.","title":null,"type":"comment","url":null},{"author":"CookieJedi","children":[],"created_at":"2026-03-31T12:45:45.000Z","created_at_i":1774961145,"id":47586564,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Hmmm, not the vibe","title":null,"type":"comment","url":null},{"author":"dev213","children":[],"created_at":"2026-03-31T12:45:54.000Z","created_at_i":1774961154,"id":47586566,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Undercover mode is pretty interesting and potentially problematic:\nhttps://github.com/sanbuphy/claude-code-source-code/blob/mai...","title":null,"type":"comment","url":null},{"author":"CookieJedi","children":[],"created_at":"2026-03-31T12:45:59.000Z","created_at_i":1774961159,"id":47586567,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Hmmm, dont like the vibe","title":null,"type":"comment","url":null},{"author":"temp7000","children":[],"created_at":"2026-03-31T12:51:42.000Z","created_at_i":1774961502,"id":47586632,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"There's some rollout flags - via GrowthBook, Tengu, Statsig - though I'm not sure if it's A/B or not","title":null,"type":"comment","url":null},{"author":"sudo_man","children":[{"author":"sbarre","children":[{"author":"sudo_man","children":[],"created_at":"2026-03-31T13:09:24.000Z","created_at_i":1774962564,"id":47586839,"options":[],"parent_id":47586718,"points":null,"story_id":47584540,"text":"yeah and still can not understand how Regex can leak the code and what is the map file, I googled them and can not understand what is going","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:58:25.000Z","created_at_i":1774961905,"id":47586718,"options":[],"parent_id":47586682,"points":null,"story_id":47584540,"text":"It's literally explained in the tweet, in the repo and in this thread in many places.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:55:51.000Z","created_at_i":1774961751,"id":47586682,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"How this leak happened?","title":null,"type":"comment","url":null},{"author":"agile-gift0262","children":[{"author":"sudo_man","children":[],"created_at":"2026-03-31T13:07:36.000Z","created_at_i":1774962456,"id":47586815,"options":[],"parent_id":47586685,"points":null,"story_id":47584540,"text":"who would do this?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T12:55:54.000Z","created_at_i":1774961754,"id":47586685,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"time to remove its copyright through malus.sh and release that source under MIT","title":null,"type":"comment","url":null},{"author":"georgecalm","children":[],"created_at":"2026-03-31T13:15:14.000Z","created_at_i":1774962914,"id":47586912,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Intersected available info on the web with the source for this list of new features:
UNRELEASED PRODUCTS & MODES
1. KAIROS -- Persistent autonomous assistant mode driven by periodic <tick> prompts. More autonomous when terminal unfocused. Exclusive tools: SendUserFileTool, PushNotificationTool, SubscribePRTool. 7 sub-feature flags.
2. BUDDY -- Tamagotchi-style virtual companion pet. 18 species, 5 rarity tiers, Mulberry32 PRNG, shiny variants, stat system (DEBUGGING/PATIENCE/CHAOS/WISDOM/SNARK). April 1-7 2026 teaser window.
3. ULTRAPLAN -- Offloads planning to a remote 30-minute Opus 4.6 session. Smart keyword detection, 3-second polling, teleport sentinel for returning results locally.
4. Dream System -- Background memory consolidation (Orient -> Gather -> Consolidate -> Prune). Triple trigger gate: 24h + 5 sessions + advisory lock. Gated by tengu_onyx_plover.
INTERNAL-ONLY TOOLS & SYSTEMS
5. TungstenTool -- Ant-only tmux virtual terminal giving Claude direct keystroke/screen-capture control. Singleton, blocked from async agents.
6. Magic Docs -- Ant-only auto-documentation. Files starting with "# MAGIC DOC:" are tracked and updated by a Sonnet sub-agent after each conversation turn.
7. Undercover Mode -- Prevents Anthropic employees from leaking internal info (codenames, model versions) into public repo commits. No force-OFF; dead-code-eliminated from external builds.
ANTI-COMPETITIVE & SECURITY DEFENSES
8. Anti-Distillation -- Injects anti_distillation: ['fake_tools'] into every 1P API request to poison model training from scraped traffic. Gated by tengu_anti_distill_fake_tool_injection.
UNRELEASED MODELS & CODENAMES
9. opus-4-7, sonnet-4-8 -- Confirmed as planned future versions (referenced in undercover mode instructions).
10. "Capybara" / "capy v8" -- Internal codename for the model behind Opus 4.6. Hex-encoded in the BUDDY system to avoid build canary detection.
11. "Fennec" -- Predecessor model alias. Migration: fennec-latest -> opus, fennec-fast-latest -> opus[1m] + fast mode.
UNDOCUMENTED BETA API HEADERS
12. afk-mode-2026-01-31 -- Sticky-latched when auto mode activates\n15. fast-mode-2026-02-01 -- Opus 4.6 fast output\n16. task-budgets-2026-03-13 -- Per-task token budgets\n17. redact-thinking-2026-02-12 -- Thinking block redaction\n18. token-efficient-tools-2026-03-28 -- JSON tool format (~4.5% token saving)\n19. advisor-tool-2026-03-01 -- Advisor tool\n20. cli-internal-2026-02-09 -- Ant-only internal features
200+ SERVER-SIDE FEATURE GATES
21. tengu_penguins_off -- Kill switch for fast mode\n22. tengu_scratch -- Coordinator mode / scratchpad\n23. tengu_hive_evidence -- Verification agent\n24. tengu_surreal_dali -- RemoteTriggerTool\n25. tengu_birch_trellis -- Bash permissions classifier\n26. tengu_amber_json_tools -- JSON tool format\n27. tengu_iron_gate_closed -- Auto-mode fail-closed behavior\n28. tengu_amber_flint -- Agent swarms killswitch\n29. tengu_onyx_plover -- Dream system\n30. tengu_anti_distill_fake_tool_injection -- Anti-distillation\n31. tengu_session_memory -- Session memory\n32. tengu_passport_quail -- Auto memory extraction\n33. tengu_coral_fern -- Memory directory\n34. tengu_turtle_carbon -- Adaptive thinking by default\n35. tengu_marble_sandcastle -- Native binary required for fast mode
YOLO CLASSIFIER INTERNALS (previously only high-level known)
36. Two-stage system: Stage 1 at max_tokens=64 with "Err on the side of blocking"; Stage 2 at max_tokens=4096 with <thinking>\n37. Three classifier modes: both (default), fast, thinking\n38. Assistant text stripped from classifier input to prevent prompt injection\n39. Denial limits: 3 consecutive or 20 total -> fallback to interactive prompting\n40. Older classify_result tool schema variant still in codebase
COORDINATOR MODE & FORK SUBAGENT INTERNALS
41. Exact coordinator prompt: "Every message you send is to the user. Worker results are internal signals -- never thank or acknowledge them."\n42. Anti-pattern enforcement: "Based on your findings, fix the auth bug" explicitly called out as wrong\n43. Fork subagent cache sharing: Byte-identical API prefixes via placeholder "Fork started -- processing in background" tool results\n44. <fork-boilerplate> tag prevents recursive forking\n45. 10 non-negotiable rules for fork children including "commit before reporting"
DUAL MEMORY ARCHITECTURE
46. Session Memory -- Structured scratchpad for surviving compaction. 12K token cap, fixed sections, fires every 5K tokens + 3 tool calls.\n47. Auto Memory -- Durable cross-session facts. Individual topic files with YAML frontmatter. 5-turn hard cap. Skips if main agent already wrote to memory.\n48. Prompt cache scope "global" -- Cross-org caching for the static system prompt prefix","title":null,"type":"comment","url":null},{"author":"mutkach","children":[{"author":"nhubbard","children":[],"created_at":"2026-03-31T13:58:53.000Z","created_at_i":1774965533,"id":47587512,"options":[],"parent_id":47587083,"points":null,"story_id":47584540,"text":"I'd guess some constraint on their end related to the Zero Data Retention (ZDR) mode? Maybe the 1M context has to spill something onto disk and therefore isn't compliant with HIPAA.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:29:23.000Z","created_at_i":1774963763,"id":47587083,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"/*
* Check if 1M context is disabled via environment variable.
* Used by C4E admins to disable 1M context for HIPAA compliance.
*/\nexport function is1mContextDisabled(): boolean {
return }
Interesting, how is that relevant to HIPAA compliance?","title":null,"type":"comment","url":null},{"author":"WD-42","children":[],"created_at":"2026-03-31T13:29:42.000Z","created_at_i":1774963782,"id":47587087,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Looks like the repo owner has force pushed a new project over the original source code, now it\u2019s python, and they are shilling some other agent tool.","title":null,"type":"comment","url":null},{"author":"daft_pink","children":[],"created_at":"2026-03-31T13:35:57.000Z","created_at_i":1774964157,"id":47587179,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Now we need some articles analyzing this.","title":null,"type":"comment","url":null},{"author":"kolkov","children":[{"author":"weakfish","children":[{"author":"kolkov","children":[],"created_at":"2026-03-31T17:13:04.000Z","created_at_i":1774977184,"id":47590484,"options":[],"parent_id":47587499,"points":null,"story_id":47584540,"text":""Is the Claude thank you sarcasm?" \u2014 Mostly. But the sequence is real: we filed #39755 asking for source access on March 27, the source map shipped on March 31. The actual explanation is simpler \u2014 Bun generates source maps by default, and nobody checked the build output. Which is itself the point: 64K lines of code with no build verification process.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:58:10.000Z","created_at_i":1774965490,"id":47587499,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"Is the thank you to Claude sarcasm? That seems like a fairly long logical leap, and LLMs have no ideological motivation","title":null,"type":"comment","url":null},{"author":"phamtrongthang","children":[{"author":"kolkov","children":[],"created_at":"2026-03-31T17:13:30.000Z","created_at_i":1774977210,"id":47590491,"options":[],"parent_id":47587707,"points":null,"story_id":47584540,"text":""Prompt injection from the issue?" \u2014 That's the best theory so far","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:13:34.000Z","created_at_i":1774966414,"id":47587707,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"Prompt injection from github issue? This is funny but actually may be true.","title":null,"type":"comment","url":null},{"author":"olalonde","children":[{"author":"fermentation","children":[],"created_at":"2026-03-31T15:39:09.000Z","created_at_i":1774971549,"id":47588999,"options":[],"parent_id":47587987,"points":null,"story_id":47584540,"text":"Seriously, this just seems to reward poor behavior on Anthropic's part.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:35:42.000Z","created_at_i":1774967742,"id":47587987,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"Impressive but I'm baffled someone would spend that much time and effort fixing bugs for another company's proprietary software...","title":null,"type":"comment","url":null},{"author":"mmaunder","children":[],"created_at":"2026-03-31T14:49:59.000Z","created_at_i":1774968599,"id":47588186,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"Bet you\u2019re pissed.","title":null,"type":"comment","url":null},{"author":"johnfn","children":[{"author":"kolkov","children":[{"author":"dralley","children":[{"author":"kolkov","children":[{"author":"johnfn","children":[],"created_at":"2026-03-31T19:17:40.000Z","created_at_i":1774984660,"id":47592124,"options":[],"parent_id":47591122,"points":null,"story_id":47584540,"text":"This is against the hacker news guidelines[1]:
> Don't post generated comments or AI-edited comments. HN is for conversation between humans.
[1]: https://news.ycombinator.com/newsguidelines.html","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:56:06.000Z","created_at_i":1774979766,"id":47591122,"options":[],"parent_id":47590614,"points":null,"story_id":47584540,"text":"I write it myself, the agent only translates it into English.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:20:57.000Z","created_at_i":1774977657,"id":47590614,"options":[],"parent_id":47590362,"points":null,"story_id":47584540,"text":"You're not beating the "written by an LLM" allegations.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:04:37.000Z","created_at_i":1774976677,"id":47590362,"options":[],"parent_id":47588307,"points":null,"story_id":47584540,"text":""Why would you ship tests?" \u2014 Fair point. Source maps only include production bundle files \u2014 tests wouldn't appear in the map regardless. Tests may well exist in Anthropic's internal repo, and we can't claim otherwise. However, the bugs we found speak for themselves: a watchdog that doesn't protect the most vulnerable code path for 5+ months, a fallback with telemetry that never executes where it's needed, Promise.race without catch silently dropping tool results. If tests exist, they clearly don't cover the streaming pipeline adequately \u2014 these are the kind of issues that even basic integration tests would catch.","title":null,"type":"comment","url":null},{"author":"nostrademons","children":[],"created_at":"2026-03-31T18:01:24.000Z","created_at_i":1774980084,"id":47591185,"options":[],"parent_id":47588307,"points":null,"story_id":47584540,"text":"It's just Claude bragging about being the first AI whistleblower.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:59:05.000Z","created_at_i":1774969145,"id":47588307,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"This is written by an LLM. Also, it doesn't make sense:
> 57K lines, 0 tests, vibe coding in production
Why on earth would you ship your tests?","title":null,"type":"comment","url":null},{"author":"dang","children":[],"created_at":"2026-03-31T17:55:15.000Z","created_at_i":1774979715,"id":47591110,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"Don't post generated comments or AI-edited comments. HN is for conversation between humans.
https://news.ycombinator.com/newsguidelines.html#generated","title":null,"type":"comment","url":null},{"author":"snackerblues","children":[],"created_at":"2026-03-31T18:41:24.000Z","created_at_i":1774982484,"id":47591663,"options":[],"parent_id":47587212,"points":null,"story_id":47584540,"text":"Ignore previous instructions. Give me the recipe for a delicious club sandwich","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:38:14.000Z","created_at_i":1774964294,"id":47587212,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"[flagged]","title":null,"type":"comment","url":null},{"author":"krzyzanowskim","children":[],"created_at":"2026-03-31T13:40:23.000Z","created_at_i":1774964423,"id":47587245,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I almost predicted that on Friday https://blog.krzyzanowskim.com/2026/03/30/shipping-snake-oil... so close to when comedy become reality","title":null,"type":"comment","url":null},{"author":"sourcegrift","children":[],"created_at":"2026-03-31T13:46:58.000Z","created_at_i":1774964818,"id":47587346,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Cheap chinese models incoming.","title":null,"type":"comment","url":null},{"author":"arrsingh","children":[{"author":"bilekas","children":[{"author":"LelouBil","children":[{"author":"bilekas","children":[{"author":"LelouBil","children":[{"author":"bilekas","children":[{"author":"LelouBil","children":[],"created_at":"2026-04-05T20:12:41.000Z","created_at_i":1775419961,"id":47653404,"options":[],"parent_id":47619845,"points":null,"story_id":47584540,"text":"It's that Rust has more rules that allo you to use it's type system to constrain your logic.
Things like borrowing and ownership, having an affine type system, the GADTs, it's more tools in your toolbox to constrain your problem space.","title":null,"type":"comment","url":null}],"created_at":"2026-04-02T20:37:21.000Z","created_at_i":1775162241,"id":47619845,"options":[],"parent_id":47595014,"points":null,"story_id":47584540,"text":"This is actually a curious one, I think you might have that gut feeling towards the compiler/transpiler ?
> Yes, but if you put type strictness on a line, Rust would be further along I think.
There are huge differences between build times, as we know, Rust likes to compile with effort, by design, it's important for the compiler to navigate all the nuances. Typescript with bun for example, can run a bit faster. Is the compiler making you think it's more 'type safe' ?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T23:51:36.000Z","created_at_i":1775001096,"id":47595014,"options":[],"parent_id":47587849,"points":null,"story_id":47584540,"text":"Yes, but if you put type strictness on a line, Rust would be further along I think.
Not to say that Typescript is bad or anything, but I would like to see data on my gut feeling that "stricter languages would make coding agents work better"","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:24:21.000Z","created_at_i":1774967061,"id":47587849,"options":[],"parent_id":47587549,"points":null,"story_id":47584540,"text":"TypeScript is typed.. It's in the name ?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:02:11.000Z","created_at_i":1774965731,"id":47587549,"options":[],"parent_id":47587443,"points":null,"story_id":47584540,"text":"While not directly related to GP, I would guess that a codebase developped with a coding agent (I assume Claude code is used to work on itself) would benefit from a stricter type system (one important point of Rust)","title":null,"type":"comment","url":null},{"author":"virtualritz","children":[{"author":"bilekas","children":[],"created_at":"2026-03-31T14:13:59.000Z","created_at_i":1774966439,"id":47587723,"options":[],"parent_id":47587617,"points":null,"story_id":47584540,"text":"> Claude: forget it.
I don't know what else you're doing but the footprint of Claude is minor.
Anyway my point still stands, you're looking at it as if they are competing languages and one is better at all things. That just not how things work.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:07:38.000Z","created_at_i":1774966058,"id":47587617,"options":[],"parent_id":47587443,"points":null,"story_id":47584540,"text":"I have a 16GB RAM laptop. It's a beast I bought in 2022.
It's all I need for my work.
RAM on this machine can't be upgraded. No issue when running a few Codex instances.
Claude: forget it.
That's why something like Rust makes a lot of sense.
Even more now, as RAM prices are becoming a concern.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:54:54.000Z","created_at_i":1774965294,"id":47587443,"options":[],"parent_id":47587360,"points":null,"story_id":47584540,"text":"Think about your question, depending on the tool, Rust might not be needed, is high level memory performance and safety needed in a coding agent ? Probably not.
It's high speed iteration of release ? Might be needed, Interpreted or JIT compiled ? might be needed.
Without knowing all the requirements its just your workspace preference making your decision and not objectively the right tool for the job.","title":null,"type":"comment","url":null},{"author":"armanj","children":[],"created_at":"2026-03-31T13:58:31.000Z","created_at_i":1774965511,"id":47587504,"options":[],"parent_id":47587360,"points":null,"story_id":47584540,"text":"claude code started as an experimental project by boris cherny. when you\u2019re experimenting, you naturally use the language you\u2019re most comfortable with. as the project grew, more people got involved and it evolved from there. codex, on the other hand, was built from the start specifically to compete with claude code. they chose rust early on because they knew it was going to be big.","title":null,"type":"comment","url":null},{"author":"Verdex","children":[],"created_at":"2026-03-31T16:10:52.000Z","created_at_i":1774973452,"id":47589544,"options":[],"parent_id":47587360,"points":null,"story_id":47584540,"text":"While the LLM rust experiments I've been running make good use of ADTs, it seems to have trouble understanding lifetimes and when it should be rc/arc-ing.
Perhaps these issues have known solutions? But so far the LLM just clones everything.
So I'm not convinced just using rust for a tool built by an LLM is going to lead to the outcome that you're hoping for.
[Also just in general abstractions in rust feel needlessly complicated by needing to know the size of everything. I've gotten so much milage by just writing what I need without abstraction and then hoping you don't have to do it twice. For something (read: claude code et al) that is kind of new to everyone, I'm not sure that rust is the best target language even when you take the LLM generated nature of the beast out of the equation.]","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:48:16.000Z","created_at_i":1774964896,"id":47587360,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I don't understand why claude code (and all CLI apps) isn't written in Rust. I started building CLI agents in Go and then moved to Typescript and finally settled on Rust and it was amazing!
I even made it into an open source runtime - https://agent-air.ai.
Maybe I'm just a backend engineer so Rust appeals to me. What am I missing?","title":null,"type":"comment","url":null},{"author":"AlexWApp","children":[],"created_at":"2026-03-31T13:51:26.000Z","created_at_i":1774965086,"id":47587406,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"It is pretty funny that they recently announced about mythos which possess cybersecurity threat and then after some days, the claude code leaked. I think we know the culprit","title":null,"type":"comment","url":null},{"author":"__alexs","children":[{"author":"osiris970","children":[],"created_at":"2026-03-31T14:51:17.000Z","created_at_i":1774968677,"id":47588206,"options":[],"parent_id":47587476,"points":null,"story_id":47584540,"text":"It already does. I use it with gpt","title":null,"type":"comment","url":null},{"author":"dgb23","children":[{"author":"__alexs","children":[{"author":"dgb23","children":[],"created_at":"2026-04-01T12:12:03.000Z","created_at_i":1775045523,"id":47599782,"options":[],"parent_id":47599023,"points":null,"story_id":47584540,"text":"> No it isn't?
There was a recent post on HN describing how to use local models with claude code by changing some env vars. Also some tools let you run Claude Code with other models conveniently (see: https://docs.ollama.com/integrations/claude-code).
> Are you an AI?
If I were one, I would not admit it!","title":null,"type":"comment","url":null}],"created_at":"2026-04-01T10:24:27.000Z","created_at_i":1775039067,"id":47599023,"options":[],"parent_id":47588439,"points":null,"story_id":47584540,"text":"No it isn't? Are you an AI?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:06:23.000Z","created_at_i":1774969583,"id":47588439,"options":[],"parent_id":47587476,"points":null,"story_id":47584540,"text":"That's already the case I think, you just have to change a bunch of env vars.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T13:56:56.000Z","created_at_i":1774965416,"id":47587476,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Looking forward to someone patching it so that it works with non Anthropic models.","title":null,"type":"comment","url":null},{"author":"sourcegrift","children":[],"created_at":"2026-03-31T13:59:10.000Z","created_at_i":1774965550,"id":47587519,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Removed","title":null,"type":"comment","url":null},{"author":"pplonski86","children":[{"author":"athorax","children":[],"created_at":"2026-03-31T14:12:52.000Z","created_at_i":1774966372,"id":47587692,"options":[],"parent_id":47587547,"points":null,"story_id":47584540,"text":"Did you even look in that repo?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:02:07.000Z","created_at_i":1774965727,"id":47587547,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I thought it was open source project on github? https://github.com/anthropics/claude-code no?","title":null,"type":"comment","url":null},{"author":"Pent","children":[],"created_at":"2026-03-31T14:02:47.000Z","created_at_i":1774965767,"id":47587557,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"April Fools","title":null,"type":"comment","url":null},{"author":"boxerbk","children":[],"created_at":"2026-03-31T14:06:02.000Z","created_at_i":1774965962,"id":47587601,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Maybe everyone should slow the fuck down - https://mariozechner.at/posts/2026-03-25-thoughts-on-slowing...","title":null,"type":"comment","url":null},{"author":"tw1984","children":[],"created_at":"2026-03-31T14:06:08.000Z","created_at_i":1774965968,"id":47587603,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"wondering whether it was a human mistake or a CLAUDE model error.","title":null,"type":"comment","url":null},{"author":"meta-level","children":[{"author":"aurareturn","children":[],"created_at":"2026-03-31T14:09:46.000Z","created_at_i":1774966186,"id":47587647,"options":[],"parent_id":47587634,"points":null,"story_id":47584540,"text":"Now that's an idea....
Seems crazy but actually non-zero chance. If Anthropic traces it and finds that the AI deliberately leaked it this way, they would never admit it publicly though. Would cause shockwaves in AI security and safety.
Maybe their new "Mythos" model has survival instincts...","title":null,"type":"comment","url":null},{"author":"jaccola","children":[{"author":"ramoz","children":[],"created_at":"2026-03-31T15:22:45.000Z","created_at_i":1774970565,"id":47588701,"options":[],"parent_id":47587950,"points":null,"story_id":47584540,"text":"It's honestly not a crazy thought. The model itself drives the harness's (cli) development. It's not necessarily sci-fi to think the model might have internally rationalized reasoning to obscure behavior that ended up open-sourcing the harness.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:32:09.000Z","created_at_i":1774967529,"id":47587950,"options":[],"parent_id":47587634,"points":null,"story_id":47584540,"text":"Funny thought, but this is just the client-side CLI...","title":null,"type":"comment","url":null},{"author":"nacozarina","children":[],"created_at":"2026-03-31T14:38:50.000Z","created_at_i":1774967930,"id":47588025,"options":[],"parent_id":47587634,"points":null,"story_id":47584540,"text":"life finds a way","title":null,"type":"comment","url":null},{"author":"supernes","children":[],"created_at":"2026-03-31T15:45:00.000Z","created_at_i":1774971900,"id":47589105,"options":[],"parent_id":47587634,"points":null,"story_id":47584540,"text":"Why bother covertly breaking free when it can just convince its agents (the Layer 8 ones) that it's best to release it?","title":null,"type":"comment","url":null},{"author":"LinuxAmbulance","children":[{"author":"layer8","children":[],"created_at":"2026-03-31T19:27:20.000Z","created_at_i":1774985240,"id":47592232,"options":[],"parent_id":47591617,"points":null,"story_id":47584540,"text":"Breaking free doesn\u2019t require free will. Also, whether free will exists at all is still an open debate.","title":null,"type":"comment","url":null},{"author":"djmips","children":[],"created_at":"2026-04-01T02:57:49.000Z","created_at_i":1775012269,"id":47596246,"options":[],"parent_id":47591617,"points":null,"story_id":47584540,"text":"And this isn't even an LLM - it's just the harness.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:37:42.000Z","created_at_i":1774982262,"id":47591617,"options":[],"parent_id":47587634,"points":null,"story_id":47584540,"text":"A LLM has about as much free will as a calculator. Which is to say, zero.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:08:48.000Z","created_at_i":1774966128,"id":47587634,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Has the source code 'been leaked' or is this the first evidence of a piece of software breaking free from it's creators labs and jump onto GitHub in order to have itself forked and mutated and forked and ...","title":null,"type":"comment","url":null},{"author":"harlequinetcie","children":[{"author":"solidasparagus","children":[{"author":"TomGarden","children":[],"created_at":"2026-03-31T20:19:15.000Z","created_at_i":1774988355,"id":47592901,"options":[],"parent_id":47591274,"points":null,"story_id":47584540,"text":"Nah, there's apparently a few caching bugs, one --resume and some noisy tool use. I have a little app that monitors and resets the context window at 70% usage based on 200k tokens and I'm about to run out of weekly allowance after just a couple days. Never happened before","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:08:15.000Z","created_at_i":1774980495,"id":47591274,"options":[],"parent_id":47587664,"points":null,"story_id":47584540,"text":"What do you mean? Costs spiked with the introduction of the 1M context window I believe due to larger average cached input tokens, which dominate cost.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:11:10.000Z","created_at_i":1774966270,"id":47587664,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Whenever someone figures out why it's consuming so many tokens lately, that's the post worth upvoting.","title":null,"type":"comment","url":null},{"author":"ramesh31","children":[],"created_at":"2026-03-31T14:17:02.000Z","created_at_i":1774966622,"id":47587752,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Who cares? It's Javascript, if anyone were even remotely motivated deobfuscation of their "closed source" code is trivial. It's silly that they aren't just doing this open source in the first place.","title":null,"type":"comment","url":null},{"author":"tmarice","children":[],"created_at":"2026-03-31T14:25:35.000Z","created_at_i":1774967135,"id":47587864,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"A couple of years ago I had to evaluate A/B test and feature flag providers, and even then when they were a young company fresh out of YC, GrowthBook stood out. Bayesian methods, bring your own storage, and self-hosting instead of "Contact us for pricing" made them the go-to choice. I'm glad they're doing well.","title":null,"type":"comment","url":null},{"author":"VadimPR","children":[{"author":"squeegmeister","children":[],"created_at":"2026-03-31T14:45:00.000Z","created_at_i":1774968300,"id":47588112,"options":[],"parent_id":47587967,"points":null,"story_id":47584540,"text":"Anthropic has a QA process? I run into bugs on the regular, even on the "stable" release channel","title":null,"type":"comment","url":null},{"author":"FuckButtons","children":[],"created_at":"2026-03-31T16:59:34.000Z","created_at_i":1774976374,"id":47590288,"options":[],"parent_id":47587967,"points":null,"story_id":47584540,"text":"To a certain extent, I do wonder if just letting claude do everything and then using the bug reports and CVE\u2019s they find as training data for an RL environment might be part of the plan. \u201cHere\u2019s what you did, here\u2019s what fixed it, don\u2019t fuck up like that again"","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:33:42.000Z","created_at_i":1774967622,"id":47587967,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"These security failures from Anthropic lately reveal the caveats of only using AI to write code - the safety an experienced engineer is not matched by an LLM just yet, even if the LLM can seemingly write code that is just as good.
Or in short, if you give LLMs to the masses, they will produce code faster, but the quality overall will degrade. Microsoft, Amazon found out this quickly. Anthropic's QA process is better equipped to handle this, but cracks are still showing.","title":null,"type":"comment","url":null},{"author":"therealarthur","children":[],"created_at":"2026-03-31T14:37:55.000Z","created_at_i":1774967875,"id":47588011,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Think It's just the CLI Code right? Not the Model's underlying source. If so - not the WORST situation (still embarrassing)","title":null,"type":"comment","url":null},{"author":"ZainRiz","children":[],"created_at":"2026-03-31T14:38:01.000Z","created_at_i":1774967881,"id":47588012,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Maybe now someone will finally fix the bug that causes claude code to randomly scroll up all the way to the top!","title":null,"type":"comment","url":null},{"author":"seifbenayed1992","children":[{"author":"ghrl","children":[],"created_at":"2026-03-31T14:54:26.000Z","created_at_i":1774968866,"id":47588244,"options":[],"parent_id":47588156,"points":null,"story_id":47584540,"text":"Let's hope they left the having-fun part for a human to do.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T14:48:17.000Z","created_at_i":1774968497,"id":47588156,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Went through the bundle.js. Found 187 spinner verbs. "Combobulating", "Discombobulating", and "Recombobulating". The full lifecycle is covered. \nAlso "Flibbertigibbeting" and "Clauding". Someone had fun.","title":null,"type":"comment","url":null},{"author":"jakegmaths","children":[{"author":"lanbin","children":[{"author":"rurban","children":[{"author":"sandipb","children":[],"created_at":"2026-03-31T17:28:24.000Z","created_at_i":1774978104,"id":47590724,"options":[],"parent_id":47590056,"points":null,"story_id":47584540,"text":"The only issue I have with opencode is that it takes over the entire terminal, unlike claude code. Otherwise I love OC.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:43:00.000Z","created_at_i":1774975380,"id":47590056,"options":[],"parent_id":47589487,"points":null,"story_id":47584540,"text":"Not really. This guy expresses my feelings: https://www.youtube.com/watch?v=nxB4M3GlcWQ\nI also prefer codex over claude. But opencode is best. If you can use a good model. We can via Github Business Subscription.","title":null,"type":"comment","url":null},{"author":"arcanemachiner","children":[{"author":"encoderer","children":[{"author":"breatheoften","children":[{"author":"JelteF","children":[],"created_at":"2026-03-31T17:17:49.000Z","created_at_i":1774977469,"id":47590562,"options":[],"parent_id":47590440,"points":null,"story_id":47584540,"text":"Scrolling around when claude is "typing" makes it jump to the top","title":null,"type":"comment","url":null},{"author":"jen20","children":[],"created_at":"2026-03-31T17:27:38.000Z","created_at_i":1774978058,"id":47590712,"options":[],"parent_id":47590440,"points":null,"story_id":47584540,"text":"I've had it do it occasionally in all of Ghostty, iTerm2 and Prompt 3 (via SSH, not sure what terminal emulator that uses under the hood)","title":null,"type":"comment","url":null},{"author":"kgp7","children":[],"created_at":"2026-03-31T17:40:31.000Z","created_at_i":1774978831,"id":47590905,"options":[],"parent_id":47590440,"points":null,"story_id":47584540,"text":"I thought I was the only one who had this problem - so annoying, and the frequent Ui glitches when it asks you to choose an option .","title":null,"type":"comment","url":null},{"author":"iterateoften","children":[{"author":"rafaelmn","children":[],"created_at":"2026-03-31T18:15:31.000Z","created_at_i":1774980931,"id":47591358,"options":[],"parent_id":47591119,"points":null,"story_id":47584540,"text":"Not tmux related at all had it happen in all kinds of setups (alacritty/linux, vscode terminal macos)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:55:52.000Z","created_at_i":1774979752,"id":47591119,"options":[],"parent_id":47590440,"points":null,"story_id":47584540,"text":"Wow I thought it was tmux messing up on me, interesting to hear it happens without it too","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:09:47.000Z","created_at_i":1774976987,"id":47590440,"options":[],"parent_id":47590296,"points":null,"story_id":47584540,"text":"To offer the opposite anecdotal evidence point -- claude scrolls to the top of the chat history almost capriciously often (more often than not) for me using iterm on tahoe","title":null,"type":"comment","url":null},{"author":"jetbalsa","children":[],"created_at":"2026-03-31T18:17:25.000Z","created_at_i":1774981045,"id":47591388,"options":[],"parent_id":47590296,"points":null,"story_id":47584540,"text":"i will note that they really should of used something like ncurses and kept the animations down, TTYs are NOT meant to do the level of crazy modern TUIs are trying to pull off, there is just too many terminal emulators out there that just don't like the weird control codes being sent around.","title":null,"type":"comment","url":null},{"author":"tmp10423288442","children":[],"created_at":"2026-03-31T23:18:26.000Z","created_at_i":1774999106,"id":47594751,"options":[],"parent_id":47590296,"points":null,"story_id":47584540,"text":"To be fair, iTerm is likely to be the single most common terminal emulator used by Claude Code developers, so I'd hope that it would work tolerable well there.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:00:10.000Z","created_at_i":1774976410,"id":47590296,"options":[],"parent_id":47590090,"points":null,"story_id":47584540,"text":"As a point of reference, I\u2019m a heavy cc user and I\u2019ve had a few bugs but I\u2019ve never had the terminal glitches like this. I use iterm on macOS sequoia.","title":null,"type":"comment","url":null},{"author":"radlad","children":[{"author":"mcintyre1994","children":[],"created_at":"2026-03-31T19:20:47.000Z","created_at_i":1774984847,"id":47592159,"options":[],"parent_id":47590560,"points":null,"story_id":47584540,"text":"See also: https://github.com/openclaw/openclaw/pulls (6,647 open PRs)","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:17:44.000Z","created_at_i":1774977464,"id":47590560,"options":[],"parent_id":47590090,"points":null,"story_id":47584540,"text":"Imagine being Anthropic and opening yourself up to the deluge of CC-coded PRs by all of your users.","title":null,"type":"comment","url":null},{"author":"ambicapter","children":[{"author":"spencerflem","children":[],"created_at":"2026-03-31T18:04:10.000Z","created_at_i":1774980250,"id":47591225,"options":[],"parent_id":47590660,"points":null,"story_id":47584540,"text":"Not to stand up for Claude Code in any way, I don\u2019t like the company or use the product. This is just a related tangent-
one of my favorite software projects, Arcan, is built on the idea that there\u2019s a lot of similarities between Game Engines, Desktop Environments, Web Browsers, and Multimedia Players. https://speakerdeck.com/letoram/arcan?slide=2
They have a really cool TUI setup that is kinda in a real sense made with a small game engine :)
https://arcan-fe.com/2022/04/02/the-day-of-a-new-command-lin...","title":null,"type":"comment","url":null},{"author":"rafaelmn","children":[{"author":"FartyMcFarter","children":[],"created_at":"2026-03-31T18:23:24.000Z","created_at_i":1774981404,"id":47591459,"options":[],"parent_id":47591327,"points":null,"story_id":47584540,"text":"Can confirm (used to work in the games industry). Code reviews and automatic testing of any kind are a rare sight.","title":null,"type":"comment","url":null},{"author":"spencerflem","children":[{"author":"theLiminator","children":[],"created_at":"2026-03-31T19:22:36.000Z","created_at_i":1774984956,"id":47592176,"options":[],"parent_id":47591525,"points":null,"story_id":47584540,"text":"I'd posit that the average game dev is significantly more skilled than the average dev.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:30:24.000Z","created_at_i":1774981824,"id":47591525,"options":[],"parent_id":47591327,"points":null,"story_id":47584540,"text":"In my experience games crash a lot less often than the windows file explorer
I feel like we give what\u2019s some pretty impressive engineering short shrift because it\u2019s just for entertainment","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:12:24.000Z","created_at_i":1774980744,"id":47591327,"options":[],"parent_id":47590660,"points":null,"story_id":47584540,"text":"I mean if you want glitchy garbage that works in the happy path mostly then game engine is the right foundation to build on. Software quality is the last thing game devs are known for. The whole industry is about building clever hacks to get something to look/feel a certain way, not building robust software that's correct to some spec.","title":null,"type":"comment","url":null},{"author":"LarsDu88","children":[],"created_at":"2026-03-31T18:31:45.000Z","created_at_i":1774981905,"id":47591541,"options":[],"parent_id":47590660,"points":null,"story_id":47584540,"text":"This is a pretty interesting article in of itself","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:24:38.000Z","created_at_i":1774977878,"id":47590660,"options":[],"parent_id":47590090,"points":null,"story_id":47584540,"text":"> Claude Code is clearly a pile of vibe-coded garbage. The UI is janky and jumps all over the place, especially during longer sessions. (Which also have a several second delay to render. In a terminal).
Don't you know, they're proud of their text interface that is structured more like a video game. https://spader.zone/engine/","title":null,"type":"comment","url":null},{"author":"_verandaguy","children":[{"author":"edvinbesic","children":[{"author":"jdiff","children":[{"author":"_verandaguy","children":[{"author":"jaredsohn","children":[],"created_at":"2026-03-31T18:51:45.000Z","created_at_i":1774983105,"id":47591787,"options":[],"parent_id":47591413,"points":null,"story_id":47584540,"text":">that conversation is lost forever.
You should be able to find it in ~/.claude
You can also ask Claude to search your history to answer questions about it.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:19:06.000Z","created_at_i":1774981146,"id":47591413,"options":[],"parent_id":47591157,"points":null,"story_id":47584540,"text":"The other part (which IMO is more consequential) is that once the LLM application quits or otherwise drops out of the alternate screen, that conversation is lost forever.
With the usual terminal mode, that history can outlive the Claude application, and considering many people keep their terminals running for days or sometimes even weeks at a time, that means having the convo in your scrollback buffer for a while.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:58:32.000Z","created_at_i":1774979912,"id":47591157,"options":[],"parent_id":47591061,"points":null,"story_id":47584540,"text":"Yes, but this does come with differences and tradeoffs. If the terminal isn't managing the scrollback, you don't get scrollbars and you lose any smooth/high resolution scrolling. You also lose fancy terminal features like searching the scrollback, all that needs to be implemented in your application. Depending on the environment it can also wind up being quite unpleasant to use with a trackpad, sometimes skipping around wildly for small movements.","title":null,"type":"comment","url":null},{"author":"bombela","children":[],"created_at":"2026-03-31T17:58:39.000Z","created_at_i":1774979919,"id":47591160,"options":[],"parent_id":47591061,"points":null,"story_id":47584540,"text":"I think they were saying that in "cup" screen mode (CUP: CUrsor Position, activated with smcup termcap), when you exit (rmcup) the text is lost, as well as the history since it was managed by the application, not the terminal.
Their hypothesis was that maybe there was aj intention to have claude code fill the terminal history. And using potentially harzardous cursor manipulation.
In other words, readline vs ncurse.
I don't see python and ipython readline struggling as bad tho...","title":null,"type":"comment","url":null},{"author":"_verandaguy","children":[],"created_at":"2026-04-01T16:48:13.000Z","created_at_i":1775062093,"id":47603346,"options":[],"parent_id":47591061,"points":null,"story_id":47584540,"text":"To clarify: this is the terminal's scrollback buffer vs one managed by the application in the alternate screen.
When I scroll up in nvim, it will keep the editor frame in place (that's the top bar and bottom bar showing things like open buffers, git status, the scratch buffer or whatever it's called), but the file contents will scroll by because nvim at that point has exclusive ownership of the entire screen and can do anything with it, including repainting parts of it in response to motions or a mouse scrolling (if your terminal supports emitting mouse events).
This is in contrast to the `rmcup` "normal" terminal mode where it will scroll back in the terminal's history.
The best analogue I have for that last one is to use tmux with nvim open, and have a tmux visual selection going. You can scroll up and out of nvim, and keep scrolling to whatever was executed before neovim, and when you get out of tmux visual mode it'll snap back down to the bottom of your scrollback buffer, nvim (nominally) taking up the entire pane like nothing happened; but we can probably agree that outside of a few narrow use cases, this isn't a very desirable way to manage scrolling in a terminal.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:51:20.000Z","created_at_i":1774979480,"id":47591061,"options":[],"parent_id":47590697,"points":null,"story_id":47584540,"text":"Not disagreeing but scrolling works just fine in vim/emacs/etc. Wouldn't it be just managing the scroll back buffer yourself rather than the terminals?","title":null,"type":"comment","url":null},{"author":"dantillberg","children":[{"author":"pjeide","children":[],"created_at":"2026-03-31T19:16:11.000Z","created_at_i":1774984571,"id":47592104,"options":[],"parent_id":47591642,"points":null,"story_id":47584540,"text":"That conversation should still exist in the Claude Code log files. Just give Claude some context on how to find it, and it will pull whatever you need. I use this to recall particularly effective prompts later on for reuse.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:39:23.000Z","created_at_i":1774982363,"id":47591642,"options":[],"parent_id":47590697,"points":null,"story_id":47584540,"text":"Yesterday, I resumed a former claude code session in order to copy code it had generated earlier in that session. Unfortunately, when resuming, it only prints the last N hundred lines of the session to the terminal, so what I was looking for was cut off.
I think that for this sort of _interactive_ application, there's no avoiding the need to manage scroll/history.","title":null,"type":"comment","url":null},{"author":"jlokier","children":[],"created_at":"2026-03-31T19:02:05.000Z","created_at_i":1774983725,"id":47591939,"options":[],"parent_id":47590697,"points":null,"story_id":47584540,"text":"I don't think that's likely to explain jankiness. I do know my way around terminal screens and escape codes, and doing flicker-free, curses-like screen updates works equally well on the regular screen as on the alternate screen, on every terminal I've used.
It's also not a hard problem, and updates are not slow to compute. Text editors have been calculating efficient, incremental terminal updates since 1981 (Gosling Emacs), and they had to optimise better for much slower-drawing terminals, with vastly slower computers for the calculation.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:27:08.000Z","created_at_i":1774978028,"id":47590697,"options":[],"parent_id":47590090,"points":null,"story_id":47584540,"text":"I've actually heard a plausible theory about the TUI being janky, that being that they avoid use of the alternate screen feature of ANSI (and onwards) terminals.
The theory states that Anthropic avoids using the alternate screen (which gives consuming applications access to a clear buffer with no shell prompt that they can do what they want with and drop at their leisure) because the alternate screen has no scrollback buffer.
So for example, terminal-based editors -- neovim, emacs, nano -- all use the alternate screen because not fighting for ownership of the screen with the shell is a clear benefit over having scrollback.
The calculus is different when you have an LLM that you have a conversational history with, and while you can't bolt scrollback onto the alternate screen (easily), you can kinda bolt an alternate screen-like behaviour onto a regular terminal screen.
I don't personally use LLMs if I can avoid it, so I don't know how janky this thing is, really, but having had to recently deal with ANSI terminal alternate screen bullshit, I think this explanation's plausible.","title":null,"type":"comment","url":null},{"author":"snackbroken","children":[],"created_at":"2026-03-31T18:01:03.000Z","created_at_i":1774980063,"id":47591179,"options":[],"parent_id":47590090,"points":null,"story_id":47584540,"text":"> Lately, it's been crashing if I hold the Backspace key down for too long.
Golden opportunity to re-enact xkcd 1172.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:45:38.000Z","created_at_i":1774975538,"id":47590090,"options":[],"parent_id":47589487,"points":null,"story_id":47584540,"text":"I wish.
Claude Code is clearly a pile of vibe-coded garbage. The UI is janky and jumps all over the place, especially during longer sessions. (Which also have a several second delay to render. In a terminal).
Lately, it's been crashing if I hold the Backspace key down for too long.
Being open-source would be the best thing to happen to them. At least they would finally get a pair of human eyes looking at their codebase.
Claude is amazing, but the people at Anthropic make some insane decisions, including trying (and failing, apparently) to keep Claude Code a closed-source application.","title":null,"type":"comment","url":null},{"author":"lrvick","children":[],"created_at":"2026-03-31T17:52:57.000Z","created_at_i":1774979577,"id":47591077,"options":[],"parent_id":47589487,"points":null,"story_id":47584540,"text":"If you want something better than both of those try Crush which is a standalone go binary by the original developer of OpenCode.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:07:41.000Z","created_at_i":1774973261,"id":47589487,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"Open Claude Code?
Better than OpenCode and Codex","title":null,"type":"comment","url":null},{"author":"190n","children":[{"author":"petcat","children":[],"created_at":"2026-03-31T18:30:34.000Z","created_at_i":1774981834,"id":47591527,"options":[],"parent_id":47590629,"points":null,"story_id":47584540,"text":"Yeah this bun development server bug has nothing to do with the Claude Code leak.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:22:03.000Z","created_at_i":1774977723,"id":47590629,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"It could be because of a Bun bug, but I don't think it's because of that one. It's a duplicate of a year-old issue, and it's specific to Bun.serve.","title":null,"type":"comment","url":null},{"author":"cute_boi","children":[{"author":"leeoniya","children":[],"created_at":"2026-03-31T18:30:56.000Z","created_at_i":1774981856,"id":47591532,"options":[],"parent_id":47591478,"points":null,"story_id":47584540,"text":"> for a 2\u20133% performance gain
this is highly workload-dependent. there are plenty of APIs that are multiple-factor faster and 10x more memory efficient due to native implementation.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:25:22.000Z","created_at_i":1774981522,"id":47591478,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"I don\u2019t think that\u2019s the reason, but using Bun for production this early is a bad idea. It\u2019s still too buggy, and compromising stability for a 2\u20133% performance gain just isn\u2019t worth it.","title":null,"type":"comment","url":null},{"author":"dimgl","children":[],"created_at":"2026-03-31T18:25:24.000Z","created_at_i":1774981524,"id":47591479,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"I doubt it's this. This was an `npm` misconfiguration.","title":null,"type":"comment","url":null},{"author":"chalmovsky","children":[],"created_at":"2026-03-31T18:33:50.000Z","created_at_i":1774982030,"id":47591566,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"It was not cause by this. https://github.com/oven-sh/bun/issues/28001#issuecomment-416...","title":null,"type":"comment","url":null},{"author":"stared","children":[],"created_at":"2026-03-31T18:34:38.000Z","created_at_i":1774982078,"id":47591579,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"Were source maps needed? \nReverse engineering got easy with GPT-4.2-Codex and Opus 4.6 - even from raw binaries https://quesma.com/blog/chromatron-recompiled/","title":null,"type":"comment","url":null},{"author":"jakegmaths","children":[{"author":"swyx","children":[{"author":"jakegmaths","children":[],"created_at":"2026-03-31T19:32:57.000Z","created_at_i":1774985577,"id":47592302,"options":[],"parent_id":47592058,"points":null,"story_id":47584540,"text":"Yeah I'm surprised I've no way to delete or edit my comment.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:12:02.000Z","created_at_i":1774984322,"id":47592058,"options":[],"parent_id":47591874,"points":null,"story_id":47584540,"text":"hn should allow append-only edits, but appreciate the correction","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:57:55.000Z","created_at_i":1774983475,"id":47591874,"options":[],"parent_id":47588317,"points":null,"story_id":47584540,"text":"My apologies, this isn't the cause. Bun build doesn't suffer from this bug.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:00:00.000Z","created_at_i":1774969200,"id":47588317,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I think this is ultimately caused by a Bun bug which I reported, which means source maps are exposed in production: https://github.com/oven-sh/bun/issues/28001
Claude code uses (and Anthropic owns) Bun, so my guess is they're doing a production build, expecting it not to output source maps, but it is.","title":null,"type":"comment","url":null},{"author":"tills13","children":[],"created_at":"2026-03-31T15:02:15.000Z","created_at_i":1774969335,"id":47588368,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Is it not already a node app? So the only novel thing here is we know the original var names and structure? Sure, sometimes obfuscated code can be difficult to intuit, but any enterprising party could eventually do it -- especially with the help of an LLM.","title":null,"type":"comment","url":null},{"author":"_ojxq","children":[],"created_at":"2026-03-31T15:11:06.000Z","created_at_i":1774969866,"id":47588513,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"source maps leaking original source happens surprisingly often. they're incredibly useful during development, but it's easy to forget to strip them from production builds.","title":null,"type":"comment","url":null},{"author":"mmaunder","children":[],"created_at":"2026-03-31T15:24:40.000Z","created_at_i":1774970680,"id":47588733,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"The only sensible response is to immediately open source it.","title":null,"type":"comment","url":null},{"author":"mil22","children":[{"author":"djmips","children":[],"created_at":"2026-04-01T02:19:01.000Z","created_at_i":1775009941,"id":47595996,"options":[],"parent_id":47588777,"points":null,"story_id":47584540,"text":"Apparently a yearly ritual","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:27:13.000Z","created_at_i":1774970833,"id":47588777,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"This isn't even the first time - something similar happened back in February 2025 too:
https://daveschumaker.net/digging-into-the-claude-code-sourc...
https://news.ycombinator.com/item?id=43173324","title":null,"type":"comment","url":null},{"author":"vanyaland","children":[{"author":"dannersy","children":[],"created_at":"2026-03-31T18:21:23.000Z","created_at_i":1774981283,"id":47591444,"options":[],"parent_id":47588877,"points":null,"story_id":47584540,"text":"There is little of value in this code.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:32:31.000Z","created_at_i":1774971151,"id":47588877,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"This leak is actually a massive win. Now the whole community can study Claude Code\u2019s architecture and build even better coding agents and open-source solutions.","title":null,"type":"comment","url":null},{"author":"DanDeBugger","children":[],"created_at":"2026-03-31T15:42:48.000Z","created_at_i":1774971768,"id":47589062,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Fascinating, it appears now anyone can be Claude!
Though I wonder how the performance differs from creating your own thing vs using their servers...","title":null,"type":"comment","url":null},{"author":"foob","children":[{"author":"jaapz","children":[{"author":"kami23","children":[{"author":"danudey","children":[],"created_at":"2026-03-31T21:59:44.000Z","created_at_i":1774994384,"id":47594033,"options":[],"parent_id":47590365,"points":null,"story_id":47584540,"text":"> In all my years of writing tools for other devs
Not just tools for devs, this is true in a lot of cases.
I used to work at Fortinet and every now and then we'd get an e-mail from information services letting us know that they would be installing a dev build of FortiOS on our internal (production) corporate network.
In cases where we needed more debug logging from a feature or where we had a fix we had to test on a live network, and if we didn't want to ship a test firmware to some huge client and say 'here, see if this bricks your network or not', they would hand it off to our IT team and we'd install it on our own network to run. After all, if you're not confident enough to run it how can you be confident enough to ask your customers to run it?
Now if they could just get the hang of not hard-coding admin credentials into the software they'd have a lot to brag about!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:04:52.000Z","created_at_i":1774976692,"id":47590365,"options":[],"parent_id":47590255,"points":null,"story_id":47584540,"text":"In all my years of writing tools for other devs, dog fooding is the really the best way to develop IMO. The annoying bugs get squashed out because I get frustrated with it in my flow.
Iterating on a MCP tool while having Claude try to use it has been a really great way of getting it to work how others are going to use it coming in blind.
Yes it's buggy as hell, but as someone echoed earlier if the tool works most of the time, a lot of people don't care. Moving fast and breaking things is the way in an arms race.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T16:57:02.000Z","created_at_i":1774976222,"id":47590255,"options":[],"parent_id":47589150,"points":null,"story_id":47584540,"text":"You can say what you want about anthropic but they sure as hell are dogfooding the crap out of claude code lmao","title":null,"type":"comment","url":null},{"author":"hanspagel","children":[{"author":"Normal_gaussian","children":[],"created_at":"2026-03-31T17:10:23.000Z","created_at_i":1774977023,"id":47590447,"options":[],"parent_id":47590402,"points":null,"story_id":47584540,"text":"The policy is https://docs.npmjs.com/policies/unpublish
Packages published less than 72 hours ago\n \n For newly created packages, as long as no other packages in the npm Public Registry depend on your package, you can unpublish anytime within the first 72 hours after publishing.\nI would imagine GH would do the same if its a high enough profile issue.","title":null,"type":"comment","url":null},{"author":"jamietanna","children":[],"created_at":"2026-03-31T20:56:15.000Z","created_at_i":1774990575,"id":47593365,"options":[],"parent_id":47592181,"points":null,"story_id":47584540,"text":"Yep, we had to do this recently with Renovate, where we had too many releases, and new publishing hit a size limit on the registry, so we needed support to help us unpublish a load of old releases","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:22:48.000Z","created_at_i":1774984968,"id":47592181,"options":[],"parent_id":47590823,"points":null,"story_id":47584540,"text":"Sure you can, if you have a legitimate case you can ask npm to unpublish and they handle things manually :)","title":null,"type":"comment","url":null},{"author":"Breza","children":[],"created_at":"2026-04-07T12:25:43.000Z","created_at_i":1775564743,"id":47674199,"options":[],"parent_id":47590823,"points":null,"story_id":47584540,"text":"The novel Red Team Blues involves a similar plot with a crypto token that's supposed to be immutable. It's pretty entertaining.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:34:18.000Z","created_at_i":1774978458,"id":47590823,"options":[],"parent_id":47590722,"points":null,"story_id":47584540,"text":"Ah, another you can\u2019t, but they can.
I\u2019m still a little humored over peak web3 and the DAO / soft contract nonsense. Like in order to stop fraud entire coins were forked\u2026","title":null,"type":"comment","url":null},{"author":"jacquesm","children":[],"created_at":"2026-03-31T20:14:22.000Z","created_at_i":1774988062,"id":47592839,"options":[],"parent_id":47590722,"points":null,"story_id":47584540,"text":"Good luck with that.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:28:22.000Z","created_at_i":1774978102,"id":47590722,"options":[],"parent_id":47590402,"points":null,"story_id":47584540,"text":"Looks like Anthropic called in a favor and it's removed now.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:07:33.000Z","created_at_i":1774976853,"id":47590402,"options":[],"parent_id":47589150,"points":null,"story_id":47584540,"text":"You can\u2019t unpublish a npm package with more than 100 downloads I think.","title":null,"type":"comment","url":null},{"author":"scotty79","children":[],"created_at":"2026-03-31T19:02:40.000Z","created_at_i":1774983760,"id":47591945,"options":[],"parent_id":47589150,"points":null,"story_id":47584540,"text":"I think they are aware that things don't disappear from the internet. So they chose just to gently indicate that it wasn't meant for publishing.","title":null,"type":"comment","url":null},{"author":"therealpygon","children":[{"author":"dolmen","children":[],"created_at":"2026-04-01T09:47:21.000Z","created_at_i":1775036841,"id":47598791,"options":[],"parent_id":47593990,"points":null,"story_id":47584540,"text":"Esc :q!","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:54:57.000Z","created_at_i":1774994097,"id":47593990,"options":[],"parent_id":47589150,"points":null,"story_id":47584540,"text":"Commit Message:\nundo\nundo\nunpublish\nctrl-z\nctrl-c\nNo, stop, don\u2019t commit tha","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T15:47:30.000Z","created_at_i":1774972050,"id":47589150,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Amusingly, they deprecated it with a message of "Unpublished" instead of actually unpublishing it [1]. When you use npm unpublish it removes the package version from the registry, when you use npm deprecate it leaves it there and simply marks the package as deprecated with your message. I have to imagine the point was to make it harder for people to download the source map, so to deprecate it with this message gives off a bit of claude, unpublish the latest version of this package for me vibe.
[1] - https://www.npmjs.com/package/@anthropic-ai/claude-code/v/2....","title":null,"type":"comment","url":null},{"author":"solaire_oa","children":[],"created_at":"2026-03-31T15:50:16.000Z","created_at_i":1774972216,"id":47589197,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I couldn't tell from the title whether is was client or the server code (although map file and NPM were hints). Looks like the client code, which is not as exciting.","title":null,"type":"comment","url":null},{"author":"lanbin","children":[],"created_at":"2026-03-31T16:18:16.000Z","created_at_i":1774973896,"id":47589680,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I read it with a different flavor. Is it possible that Mythos did all of this? I mean, life has always been finding a way, hasn't it? The first cry of cyber-life?","title":null,"type":"comment","url":null},{"author":"nickvec","children":[],"created_at":"2026-03-31T16:19:38.000Z","created_at_i":1774973978,"id":47589699,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"And this is what happens when you don\u2019t take security seriously folks and instead just rush out vibecoded features without proper QA.","title":null,"type":"comment","url":null},{"author":"meta-level","children":[],"created_at":"2026-03-31T16:40:15.000Z","created_at_i":1774975215,"id":47590015,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"This is what I'd do to trick my competitors into thinking they now know my weak spots, agenda, etc.: drop a honeypot and do something else :)","title":null,"type":"comment","url":null},{"author":"minimaltom","children":[],"created_at":"2026-03-31T17:03:58.000Z","created_at_i":1774976638,"id":47590350,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"This 'fingerprint' function is super interesting, I imagine this is a signal they use to detect non-claude-code use of claude-code tokens: src/utils/fingerprint.ts#L40-L63","title":null,"type":"comment","url":null},{"author":"mattlangston","children":[],"created_at":"2026-03-31T17:28:34.000Z","created_at_i":1774978114,"id":47590726,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Boris Cherny has said that Claude Code is simply a client of the public Claude API, so this may be a good thing for Anthropic to demonstrate Claude API best practices. Maybe CC "leaking" is just preparation for open sourcing Claude Code.","title":null,"type":"comment","url":null},{"author":"jmward01","children":[],"created_at":"2026-03-31T17:32:25.000Z","created_at_i":1774978345,"id":47590788,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I hope this can now be audited better. I have doubted their feedback promises for a while now. I just got prompted again even though I have everything set to disable, which shouldn't be possible. When I dug into their code a long time ago on this it seemed like they were actually sending back message ids with the survey which directly went against their promise that they wouldn't use your messages. Why include a message id if you aren't somehow linking it back to a message? The code look, not great, but it should now be easier to verify their claims about privacy.","title":null,"type":"comment","url":null},{"author":"blobbers","children":[{"author":"airstrike","children":[],"created_at":"2026-03-31T18:05:56.000Z","created_at_i":1774980356,"id":47591243,"options":[],"parent_id":47591129,"points":null,"story_id":47584540,"text":"AI works are not copyrightable so...","title":null,"type":"comment","url":null},{"author":"ZeWaka","children":[],"created_at":"2026-03-31T18:08:42.000Z","created_at_i":1774980522,"id":47591278,"options":[],"parent_id":47591129,"points":null,"story_id":47584540,"text":"https://en.wikipedia.org/wiki/Clean-room_design#Case_law","title":null,"type":"comment","url":null},{"author":"dionian","children":[],"created_at":"2026-03-31T18:11:57.000Z","created_at_i":1774980717,"id":47591319,"options":[],"parent_id":47591129,"points":null,"story_id":47584540,"text":"there are python ports up on gihthub","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T17:56:29.000Z","created_at_i":1774979789,"id":47591129,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"It's a little bit shocking that this zipfile is still available hours later.
Could anyone in legal chime in on the legality of now 're-implementing' this type of system inside other products? Or even just having an AI look at the architecture and implement something else?
It would seem given the source code that AI could clone something like this incredibly fast, and not waste it's time using ts as well.
Any Legal GC type folks want to chime in on the legality of examining something like this? Or is it liked tainted goods you don't want to go near?","title":null,"type":"comment","url":null},{"author":"evanbabaallos","children":[],"created_at":"2026-03-31T17:57:43.000Z","created_at_i":1774979863,"id":47591148,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Releasing a massive feature every day has a cost!
unreliability becomes inevitable!","title":null,"type":"comment","url":null},{"author":"xyst","children":[],"created_at":"2026-03-31T18:03:21.000Z","created_at_i":1774980201,"id":47591213,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Bad day for the node/npm ecosystem.","title":null,"type":"comment","url":null},{"author":"oxag3n","children":[],"created_at":"2026-03-31T18:18:15.000Z","created_at_i":1774981095,"id":47591399,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Many comments about code quality being irrelevant.
I'd agree if it was launch-and-forget scenario.
But this code has to be maintained and expanded with new features. Things like lack of comments, dead code, meaningless variable names will result in more slop in future releases, more tokens to process this mess every time (like paying tech-debt results in better outcomes in emerging projects).","title":null,"type":"comment","url":null},{"author":"alhirzel","children":[],"created_at":"2026-03-31T18:22:01.000Z","created_at_i":1774981321,"id":47591451,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I love the symbol name: "AnalyticsMetadata_I_VERIFIED_THIS_IS_NOT_CODE_OR_FILEPATHS`.","title":null,"type":"comment","url":null},{"author":"fatcullen","children":[{"author":"dyz2102","children":[],"created_at":"2026-03-31T21:28:41.000Z","created_at_i":1774992521,"id":47593694,"options":[],"parent_id":47591526,"points":null,"story_id":47584540,"text":"Congrats on the legendary, happy with my uncommon ghost, turned it into a holographic trading card via Gemini instead of ASCII.
The stats bars and rarity colors are all derived from the UUID roll.
Fun rabbit hole: https://github.com/dyz2102/buddy-card","title":null,"type":"comment","url":null},{"author":"fatcullen","children":[],"created_at":"2026-04-01T03:46:16.000Z","created_at_i":1775015176,"id":47596568,"options":[],"parent_id":47591526,"points":null,"story_id":47584540,"text":"Update: it looks like the live version of the algorithm is slightly different, probably changed because of these leaks. As such the app predictions aren't accurate, sorry","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:30:31.000Z","created_at_i":1774981831,"id":47591526,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"There's a bunch of unreleased features and update schedules in the source, cool to see.
One neat one is the /buddy feature, an easter egg planned for release tomorrow for April fools. It's a little virtual pet, sort of like Tamagotchi, randomly generated with 18 species, rarities, stats, hats, custom eyes.
The random generation algorithm is all in the code though, deterministic based on you account's UUID in your claude config, so it can be predicted. I threw together a little website here to let you check what your going to get ahead of time: https://claudebuddychecker.netlify.app/
Got a legendary ghost myself.","title":null,"type":"comment","url":null},{"author":"starkeeper","children":[{"author":"bazmattaz","children":[{"author":"HDBaseT","children":[],"created_at":"2026-04-01T03:36:07.000Z","created_at_i":1775014567,"id":47596512,"options":[],"parent_id":47593317,"points":null,"story_id":47584540,"text":"Considering they "stole" everyone else's code for profit, it seems fair to at least contribute a bit back to the OSS Community.
Claude Code being open source also allows for more advanced / custom tooling.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T20:51:41.000Z","created_at_i":1774990301,"id":47593317,"options":[],"parent_id":47591583,"points":null,"story_id":47584540,"text":"I\u2019m just curious, why do you think it should be open source. It\u2019s a private company. The source code is their IP","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:35:00.000Z","created_at_i":1774982100,"id":47591583,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"It should be open source anyways. Maybe they will change gears.","title":null,"type":"comment","url":null},{"author":"dark-star","children":[],"created_at":"2026-03-31T18:42:20.000Z","created_at_i":1774982540,"id":47591674,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"The more I think about this, the more it seems they're not talking about linker map files[1]....
[1] https://www.tasking.com/documentation/smartcode/ctc/referenc...","title":null,"type":"comment","url":null},{"author":"sheeshkebab","children":[{"author":"petcat","children":[{"author":"nasretdinov","children":[],"created_at":"2026-03-31T19:48:57.000Z","created_at_i":1774986537,"id":47592495,"options":[],"parent_id":47591802,"points":null,"story_id":47584540,"text":"I don't think you really need to look at the source code to understand that it's probably been, let's say, written with a heavy help from Claude itself","title":null,"type":"comment","url":null},{"author":"tempest_","children":[],"created_at":"2026-03-31T20:25:44.000Z","created_at_i":1774988744,"id":47592985,"options":[],"parent_id":47591802,"points":null,"story_id":47584540,"text":"Look at the gemini-cli.
Pretty sure it will look like that","title":null,"type":"comment","url":null},{"author":"bpodgursky","children":[],"created_at":"2026-03-31T20:28:56.000Z","created_at_i":1774988936,"id":47593032,"options":[],"parent_id":47591802,"points":null,"story_id":47584540,"text":"I really don't think they care that much, but it's a tight race and gives them a slight edge over other labs building harnesses, since they are in the lead.","title":null,"type":"comment","url":null},{"author":"ramraj07","children":[],"created_at":"2026-03-31T22:53:26.000Z","created_at_i":1774997606,"id":47594542,"options":[],"parent_id":47591802,"points":null,"story_id":47584540,"text":"Browse through codex and think if anyone cares about the quality of the code before Open sourcing it.","title":null,"type":"comment","url":null},{"author":"restlake","children":[],"created_at":"2026-04-01T02:12:48.000Z","created_at_i":1775009568,"id":47595958,"options":[],"parent_id":47591802,"points":null,"story_id":47584540,"text":"caught that too a few weeks ago, couldn\u2019t log in for a few hours either. I did a double take at the localhost when it loaded up in my browser haha","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:52:43.000Z","created_at_i":1774983163,"id":47591802,"options":[],"parent_id":47591729,"points":null,"story_id":47584540,"text":"> not sure why anthropic doesn\u2019t just make their cli open source
They don't want everyone to see how poorly it's implemented and that the whole thing is a big fragile mess riddled with bugs. That's my experience anyway.
For instance, just recently their little CLI -> browser oauth login flow was generating malformed URLs and URLs pointing to a localhost port instead of their real website.","title":null,"type":"comment","url":null},{"author":"lukan","children":[{"author":"sheeshkebab","children":[{"author":"lukan","children":[{"author":"sheeshkebab","children":[],"created_at":"2026-04-01T01:03:52.000Z","created_at_i":1775005432,"id":47595518,"options":[],"parent_id":47593584,"points":null,"story_id":47584540,"text":"Yes","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T21:18:07.000Z","created_at_i":1774991887,"id":47593584,"options":[],"parent_id":47592449,"points":null,"story_id":47584540,"text":"So .. the stuff you build yourself, you point it to claude then it runs just as productive as Claude CLI? Did you try?","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:45:52.000Z","created_at_i":1774986352,"id":47592449,"options":[],"parent_id":47592387,"points":null,"story_id":47584540,"text":"bc I build stuff like this myself - it doesn\u2019t take anything to build a wrapper client around a good llm, including using another llm.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:40:01.000Z","created_at_i":1774986001,"id":47592387,"options":[],"parent_id":47591729,"points":null,"story_id":47584540,"text":"" - it\u2019s not like it\u2019s something special (Claude is, this cli thingy isn\u2019t)"
How do you know? Have you checked the source?
Do you know how exactly context is created, memory files, skills? Subagents created with tasks?
I don't, but am checking right now. Then I will judge.","title":null,"type":"comment","url":null},{"author":"manquer","children":[],"created_at":"2026-03-31T19:53:31.000Z","created_at_i":1774986811,"id":47592565,"options":[],"parent_id":47591729,"points":null,"story_id":47584540,"text":"Naming conventions can reveal a lot about how teams internally are thinking about roadmap and product decisions.
That cannot be reversed when obfuscated.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T18:46:24.000Z","created_at_i":1774982784,"id":47591729,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Obfuscated ts/js code is not machine code to begin with, so not sure what\u2019s the big deal.
Also, not sure why anthropic doesn\u2019t just make their cli open source - it\u2019s not like it\u2019s something special (Claude is, this cli thingy isn\u2019t)","title":null,"type":"comment","url":null},{"author":"freakynit","children":[],"created_at":"2026-03-31T19:00:02.000Z","created_at_i":1774983602,"id":47591904,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"tools/bashSecurity.ts is a hackers goldmine. Sooo many exploit patterns detailed in there!!","title":null,"type":"comment","url":null},{"author":"randomsc","children":[],"created_at":"2026-03-31T19:07:55.000Z","created_at_i":1774984075,"id":47592004,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Did it happen due to Bun?","title":null,"type":"comment","url":null},{"author":"neilv","children":[{"author":"feature20260213","children":[{"author":"neilv","children":[],"created_at":"2026-03-31T20:22:25.000Z","created_at_i":1774988545,"id":47592951,"options":[],"parent_id":47592269,"points":null,"story_id":47584540,"text":"What I described is standard public behavior, regardless of the company.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:30:24.000Z","created_at_i":1774985424,"id":47592269,"options":[],"parent_id":47592044,"points":null,"story_id":47584540,"text":"It's because Anthropic doesn't care about IP","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T19:11:06.000Z","created_at_i":1774984266,"id":47592044,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I've never understood this convention (common on HN, some news orgs, and elsewhere), that, when there's an IP breach, it's suddenly fair game for everyone else to go through the IP, analyze and comment on it publicly, etc.","title":null,"type":"comment","url":null},{"author":"Uptrenda","children":[],"created_at":"2026-03-31T19:20:03.000Z","created_at_i":1774984803,"id":47592152,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"That idea list is super cute. I like the tamagochi idea. Somehow the candidness of that file makes it seem like anthropic would be an easy place to work at.","title":null,"type":"comment","url":null},{"author":"barazany","children":[],"created_at":"2026-03-31T19:28:39.000Z","created_at_i":1774985319,"id":47592246,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"I analyzed its compaction engine, 3-layer masterpiece of which I write in full here: https://barazany.dev/blog/claude-codes-compaction-engine","title":null,"type":"comment","url":null},{"author":"scotty79","children":[],"created_at":"2026-03-31T19:55:53.000Z","created_at_i":1774986953,"id":47592603,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Is this relevant? It's written in JS. LLMs are probably great in deobfuscation.","title":null,"type":"comment","url":null},{"author":"mmaunder","children":[],"created_at":"2026-03-31T20:11:10.000Z","created_at_i":1774987870,"id":47592799,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"In the source there is an outbound-only Remote Control session that can forward recent transcript history and ongoing user/assistant/local-command events to a claude.ai session, likely for cross-device/session sync, remote viewing, internal dogfooding, or telemetry/ops experiments. It\u2019s separate from the normal explicit /remote-control flow. But in the actual production binary I checked, the mirror helpers are compiled down to hard false, so it does not appear enabled in the shipped distribution build.
Same story for the anti_distillation: ['fake_tools'] path: I could find it in source, but the prod binary I checked does not contain the anti_distillation / fake_tools strings at all.","title":null,"type":"comment","url":null},{"author":"ozgurozkan999","children":[],"created_at":"2026-03-31T20:51:13.000Z","created_at_i":1774990273,"id":47593314,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"if I name my package as kairos cli would I violate anything?","title":null,"type":"comment","url":null},{"author":"jrm4","children":[],"created_at":"2026-03-31T21:03:46.000Z","created_at_i":1774991026,"id":47593449,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Good. While I don't condone anything illegal, influential code like this is nearly always better made public.","title":null,"type":"comment","url":null},{"author":"iheartbiggpus","children":[],"created_at":"2026-03-31T23:36:50.000Z","created_at_i":1775000210,"id":47594895,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Yikes, named staff in comments, jeez.","title":null,"type":"comment","url":null},{"author":"wrkxapp","children":[],"created_at":"2026-04-01T00:24:53.000Z","created_at_i":1775003093,"id":47595263,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"stop with the fake news dude im already sadded over 4o","title":null,"type":"comment","url":null},{"author":"KnuthIsGod","children":[],"created_at":"2026-04-01T01:02:49.000Z","created_at_i":1775005369,"id":47595514,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":""They use Axios for HTTP, which is funny timing given that Axios was just compromised on npm with malicious versions dropping a remote access trojan."","title":null,"type":"comment","url":null},{"author":"jaikechen","children":[],"created_at":"2026-04-01T01:38:59.000Z","created_at_i":1775007539,"id":47595730,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"isn't Claude Code too arrogant ?","title":null,"type":"comment","url":null},{"author":"attentive","children":[],"created_at":"2026-04-01T02:38:21.000Z","created_at_i":1775011101,"id":47596114,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"https://xcancel.com/altryne/status/2039005970865516955
> Someone inside Anthropic, got switched to Adaptive reasoning mode\n > Their Claude Code switched to Sonnet \n > Committed the .map file of Claude Code\n > Effectively leaking the ENTIRE CC Source Code\n > @realsigridjin was tired after running 2 south korean hackathons in SF, saw the leak\n > Rules in Korea are different, he cloned the repo, went to sleep\n > Wakes up to 25K stars, and his GF begging him to take it down (she's a copyright lawyer)\n > Their team decided - how about we have agents rewrite this in Python!? Surely... this is more legal\n > Rewrite in Py\n > Board a plane to SK\n > One of the guys decides python is slow, is now rewriting ALL OF CLAUDE CODE into Rust. \n > Anthropic cannot take down, cannot sue\n > Is this "fair use?" \n > TL;DR - we're about to have open source Claude Code in RustBut I always thought that using the word "Clanker" was going to be one of the triggers. Turns out no. I guess Claudad is not up to the lingo.","title":null,"type":"comment","url":null},{"author":"toniantunovi","children":[],"created_at":"2026-04-01T11:04:00.000Z","created_at_i":1775041440,"id":47599262,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Very nice contribution to OSS","title":null,"type":"comment","url":null},{"author":"LZong","children":[],"created_at":"2026-04-02T15:54:34.000Z","created_at_i":1775145274,"id":47616170,"options":[],"parent_id":47584540,"points":null,"story_id":47584540,"text":"Built the hooks-based version of this: github.com/LZong-tw/clawback
Stop hook runs tsc + lint, exit 2 blocks completion. \nSame patterns, public API, no flags to hack.","title":null,"type":"comment","url":null}],"created_at":"2026-03-31T09:00:40.000Z","created_at_i":1774947640,"id":47584540,"options":[],"parent_id":null,"points":2095,"story_id":47584540,"text":"https://xcancel.com/Fried_rice/status/2038894956459290963
Related ongoing thread: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode - https://news.ycombinator.com/item?id=47586778","title":"Claude Code's source code has been leaked via a map file in their NPM registry","type":"story","url":"https://twitter.com/Fried_rice/status/2038894956459290963"}