Looking back it seems like not a terrible design pattern for multiplayer turn-based games, if you can work out the kinks. Atomicity guarantees at least that there is a consistent state that won't get lost. Doing that read/write loop for an action game? Pure folly, but it's pretty funny to me."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"We ported the original Doom to SQL"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://cedardb.com/blog/sqldoom/"}},"_tags":["comment","author_noduerme","story_49948300"],"author":"noduerme","comment_text":"The game state being a SQL table just kinda triggered a memory of a year of optimization for me. One thing I'm still unsure of being a good decision or a bad one, when I wrote my casino in 2010, was having every remote call update game states on SQL tables that were used as the source of truth. With multiple players you can imagine that there would sometimes be issues. Some of the deadlock problems early on were horrific; scaling was a nightmare. But everything was atomic. No risk of lost data beyond one turn not reaching the server or deadlocking, nothing like a huge nodejs process choking on everyone's calls at the same time, or losing its memory. You always had state.
Looking back it seems like not a terrible design pattern for multiplayer turn-based games, if you can work out the kinks. Atomicity guarantees at least that there is a consistent state that won't get lost. Doing that read/write loop for an action game? Pure folly, but it's pretty funny to me.","created_at":"2026-10-05T08:09:55Z","created_at_i":1791187795,"objectID":"49962015","parent_id":49948300,"story_id":49948300,"story_title":"We ported the original Doom to SQL","story_url":"https://cedardb.com/blog/sqldoom/","updated_at":"2026-10-05T09:16:27Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"londons_explore"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"One underlying claim here is that latency within the data center will start to matter more and more for future AI compute.
That is relevant for musks in-space computation. There, latency between compute nodes, due to being hundreds of kilometres apart, cannot be low.
This pretty much limits the size of AI models which can be trained on a space datacenter to a single satellite to avoid incurring huge latency overhead both in training and inference.
With the smartest models getting bigger and bigger, that really doesn't look good for space based compute."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Homa: The end of TCP for AI clusters [video]"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=eZ8WWZzoaR0"}},"_tags":["comment","author_londons_explore","story_49957117"],"author":"londons_explore","children":[49962065],"comment_text":"One underlying claim here is that latency within the data center will start to matter more and more for future AI compute.
That is relevant for musks in-space computation. There, latency between compute nodes, due to being hundreds of kilometres apart, cannot be low.
This pretty much limits the size of AI models which can be trained on a space datacenter to a single satellite to avoid incurring huge latency overhead both in training and inference.
With the smartest models getting bigger and bigger, that really doesn't look good for space based compute.","created_at":"2026-10-05T06:54:07Z","created_at_i":1791183247,"objectID":"49961491","parent_id":49957117,"story_id":49957117,"story_title":"Homa: The end of TCP for AI clusters [video]","story_url":"https://www.youtube.com/watch?v=eZ8WWZzoaR0","updated_at":"2026-10-05T08:17:27Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"locknitpicker"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"> Why would you run encryption between nodes in an (...) cluster?
This is self-evident, isn't it?
Other than the buzzword factor, why do you think name-dropping AI makes a difference?"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Homa: The end of TCP for AI clusters [video]"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=eZ8WWZzoaR0"}},"_tags":["comment","author_locknitpicker","story_49957117"],"author":"locknitpicker","comment_text":"> Why would you run encryption between nodes in an (...) cluster?
This is self-evident, isn't it?
Other than the buzzword factor, why do you think name-dropping AI makes a difference?","created_at":"2026-10-05T06:34:20Z","created_at_i":1791182060,"objectID":"49961365","parent_id":49959762,"story_id":49957117,"story_title":"Homa: The end of TCP for AI clusters [video]","story_url":"https://www.youtube.com/watch?v=eZ8WWZzoaR0","updated_at":"2026-10-05T09:13:57Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"MrDrMcCoy"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"So that any sensitive data stupidly uploaded doesn't get leaked, which includes "zero-retention" corporate stuff. Anything that could get sniffed by employees and other agents. It really should be standard practice for any data passing between nodes, even if it's all within a site."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Homa: The end of TCP for AI clusters [video]"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=eZ8WWZzoaR0"}},"_tags":["comment","author_MrDrMcCoy","story_49957117"],"author":"MrDrMcCoy","comment_text":"So that any sensitive data stupidly uploaded doesn't get leaked, which includes "zero-retention" corporate stuff. Anything that could get sniffed by employees and other agents. It really should be standard practice for any data passing between nodes, even if it's all within a site.","created_at":"2026-10-05T02:13:48Z","created_at_i":1791166428,"objectID":"49959966","parent_id":49959762,"story_id":49957117,"story_title":"Homa: The end of TCP for AI clusters [video]","story_url":"https://www.youtube.com/watch?v=eZ8WWZzoaR0","updated_at":"2026-10-05T07:12:57Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tcdent"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"Why would you run encryption between nodes in an AI cluster?"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Homa: The end of TCP for AI clusters [video]"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=eZ8WWZzoaR0"}},"_tags":["comment","author_tcdent","story_49957117"],"author":"tcdent","children":[49959995,49960176,49959966,49961365],"comment_text":"Why would you run encryption between nodes in an AI cluster?","created_at":"2026-10-05T01:35:26Z","created_at_i":1791164126,"objectID":"49959762","parent_id":49959116,"story_id":49957117,"story_title":"Homa: The end of TCP for AI clusters [video]","story_url":"https://www.youtube.com/watch?v=eZ8WWZzoaR0","updated_at":"2026-10-05T09:18:12Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"throw0101c"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"With regards to (~6:01) "congestion control is the responsibility of the sender" and "somehow we have to get the sending nodes to stop sending so fast". Does that not exist in Ethernet/RoCE?
> Link Level Flow Control: InfiniBand uses a credit-based algorithm to guarantee lossless HCA-to-HCA communication. RoCE runs on top of Ethernet. Implementations may require lossless Ethernet network for reaching to performance characteristics similar to InfiniBand. Lossless Ethernet is typically configured via Ethernet flow control or priority flow control (PFC). Configuring a Data center bridging (DCB) Ethernet network can be more complex than configuring an InfiniBand network.[19]
* https://en.wikipedia.org/wiki/RDMA_over_Converged_Ethernet
> A sending station (computer or network switch) may be transmitting data faster than the other end of the link can accept it. Using flow control, the receiving station can signal the sender requesting suspension of transmissions until the receiver catches up. Flow control on Ethernet can be implemented at the data link layer.
* https://en.wikipedia.org/wiki/Ethernet_flow_control"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Homa: The end of TCP for AI clusters [video]"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=eZ8WWZzoaR0"}},"_tags":["comment","author_throw0101c","story_49957117"],"author":"throw0101c","children":[49958935,49958892,49958906],"comment_text":"With regards to (~6:01) "congestion control is the responsibility of the sender" and "somehow we have to get the sending nodes to stop sending so fast". Does that not exist in Ethernet/RoCE?
> Link Level Flow Control: InfiniBand uses a credit-based algorithm to guarantee lossless HCA-to-HCA communication. RoCE runs on top of Ethernet. Implementations may require lossless Ethernet network for reaching to performance characteristics similar to InfiniBand. Lossless Ethernet is typically configured via Ethernet flow control or priority flow control (PFC). Configuring a Data center bridging (DCB) Ethernet network can be more complex than configuring an InfiniBand network.[19]
* https://en.wikipedia.org/wiki/RDMA_over_Converged_Ethernet
> A sending station (computer or network switch) may be transmitting data faster than the other end of the link can accept it. Using flow control, the receiving station can signal the sender requesting suspension of transmissions until the receiver catches up. Flow control on Ethernet can be implemented at the data link layer.
* https://en.wikipedia.org/wiki/Ethernet_flow_control","created_at":"2026-10-04T22:59:39Z","created_at_i":1791154779,"objectID":"49958826","parent_id":49957117,"story_id":49957117,"story_title":"Homa: The end of TCP for AI clusters [video]","story_url":"https://www.youtube.com/watch?v=eZ8WWZzoaR0","updated_at":"2026-10-05T01:31:11Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mairogospel"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"Location:Nigeria\n Remote:Yes, also happy to be onsite\n Willing to relocate: Yes, anywhere\n Technologies: Python, C#, NodeJs, Javascript, Typescript, Docker, Terraform, SQL, Cloud Services, Backend, Data Engineering, Devops\n R\u00e9sum\u00e9/CV: https://drive.google.com/file/d/1q3yC_Q9d6N_InVMOXyyJ1ALL0ZH...\n Email: Gospelmairo@gmail.com"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Who wants to be hired? (October 2026)"}},"_tags":["comment","author_mairogospel","story_49922568"],"author":"mairogospel","comment_text":"Location:Nigeria\n Remote:Yes, also happy to be onsite\n Willing to relocate: Yes, anywhere\n Technologies: Python, C#, NodeJs, Javascript, Typescript, Docker, Terraform, SQL, Cloud Services, Backend, Data Engineering, Devops\n R\u00e9sum\u00e9/CV: https://drive.google.com/file/d/1q3yC_Q9d6N_InVMOXyyJ1ALL0ZH...\n Email: Gospelmairo@gmail.com","created_at":"2026-10-04T22:01:05Z","created_at_i":1791151265,"objectID":"49958338","parent_id":49922568,"story_id":49922568,"story_title":"Ask HN: Who wants to be hired? (October 2026)","updated_at":"2026-10-04T22:03:26Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"teo_zero"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"But Vx needs a "machine file" for each target hardware, and none is provided for Intel:
> The repository ships machine files for H100, H200, B200, A100, MI300X, Apple M4 and multi-GPU nodes
So it seems it won't work out of the box."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Vx \u2013 One Language, Every Chip"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://vxlang.org/"}},"_tags":["comment","author_teo_zero","story_49946076"],"author":"teo_zero","comment_text":"But Vx needs a "machine file" for each target hardware, and none is provided for Intel:
> The repository ships machine files for H100, H200, B200, A100, MI300X, Apple M4 and multi-GPU nodes
So it seems it won't work out of the box.","created_at":"2026-10-04T21:12:05Z","created_at_i":1791148325,"objectID":"49957931","parent_id":49951386,"story_id":49946076,"story_title":"Vx \u2013 One Language, Every Chip","story_url":"https://vxlang.org/","updated_at":"2026-10-04T21:16:10Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"spankalee"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"How would you solve that? To this day I've never seen a workable proposal.
Slots need to have a host element to select slotted nodes from. ShadowRoot provides that host. Everyone who I've seen say that it's easy comes up with an ambiguous system that doesn't work in many cases and could never be standardized.
Which is a common problem with most "web components suck" takes - they have no sympathy for standards developers or the difficulty of the problem and somehow assume the spec people are dumb or evil, and they don't do the work to show how something could realistically be better."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Why don't more developers \u201cuse the platform\u201d?"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://nolanlawson.com/2026/10/03/why-dont-more-developers-use-the-platform/"}},"_tags":["comment","author_spankalee","story_49950554"],"author":"spankalee","children":[49958541,49957298],"comment_text":"How would you solve that? To this day I've never seen a workable proposal.
Slots need to have a host element to select slotted nodes from. ShadowRoot provides that host. Everyone who I've seen say that it's easy comes up with an ambiguous system that doesn't work in many cases and could never be standardized.
Which is a common problem with most "web components suck" takes - they have no sympathy for standards developers or the difficulty of the problem and somehow assume the spec people are dumb or evil, and they don't do the work to show how something could realistically be better.","created_at":"2026-10-04T17:58:27Z","created_at_i":1791136707,"objectID":"49956217","parent_id":49954281,"story_id":49950554,"story_title":"Why don't more developers \u201cuse the platform\u201d?","story_url":"https://nolanlawson.com/2026/10/03/why-dont-more-developers-use-the-platform/","updated_at":"2026-10-05T08:50:27Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"shivang2607"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"Location: India\nRemote: Yes, anywhere\nWilling to Relocate: Yes, to Bangalore\nTechnologies: javascript, typescript, Nextjs, Nodejs, Supabase, SQL, AI Agents, Vector Databases\nResume: https://drive.google.com/file/d/1n-3chCNz61LyMl1Hju13DrzE8bA...\nEmail: shivangkh26@gmail.com
I am building and Intelligent Codebase System for AI agents and Developers: https://github.com/devlensio/devlensOSS"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Who wants to be hired? (October 2026)"}},"_tags":["comment","author_shivang2607","story_49922568"],"author":"shivang2607","comment_text":"Location: India\nRemote: Yes, anywhere\nWilling to Relocate: Yes, to Bangalore\nTechnologies: javascript, typescript, Nextjs, Nodejs, Supabase, SQL, AI Agents, Vector Databases\nResume: https://drive.google.com/file/d/1n-3chCNz61LyMl1Hju13DrzE8bA...\nEmail: shivangkh26@gmail.com
I am building and Intelligent Codebase System for AI agents and Developers: https://github.com/devlensio/devlensOSS","created_at":"2026-10-04T14:33:03Z","created_at_i":1791124383,"objectID":"49954276","parent_id":49922568,"story_id":49922568,"story_title":"Ask HN: Who wants to be hired? (October 2026)","updated_at":"2026-10-04T14:37:24Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"manu29d"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"
Location: Bengaluru, India\n Remote: Yes\n Willing to relocate: Not really, remote preferable\n Technologies: Rails, Nodejs, React, LLM\n R\u00e9sum\u00e9/CV: https://manu29d.github.io/ManuDwivedi-EngineeringManager.pdf\n Email: manu29.d@gmail.com\n Location: Bengaluru, India\n Remote: Yes\n Willing to relocate: Not really, remote preferable\n Technologies: Rails, Nodejs, React, LLM\n R\u00e9sum\u00e9/CV: https://manu29d.github.io/ManuDwivedi-EngineeringManager.pdf\n Email: manu29.d@gmail.com\nhttps://github.com/Lkruitwagen/global-fossil-fuel-supply-cha...: The resulting complex network has 6.09mn nodes and 15.70mn edges and is implemented in a graph database. With estimates of annual coal, gas, and oil demand in 13,229 global population centres and 8,165 global power stations, we use a minimum-cost flow method to estimate global asset-level energy flows.
If you want to protect your country, here are your options:
1) Invest in nuclear power and nukes (source of Uranium may be a concern). See North Korea vs Iran.
2) Solar on every structure, every vehicle is an EV. This will be the most resilient, distributed grid that can never be disrupted. Node level failures (hail on roof breaks solar) can be easily overcome. EVs are batteries on wheels! They can go charge and come back. Some on this forum may be familiar with Cattle vs Pet, Redundancy, Resiliency, Virtualization (--> Virtual Power Plants), etc -- same concepts, but applied to power generation which is designed as a monolithic circuit switched network. The idea of energy as packets, which can be Store-and-Forward (using batteries) will change everything we know about energy."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"In Ukraine, distributed renewables foil Russia's assaults"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://energytransition.org/2026/09/in-ukraine-distributed-renewables-foil-russias-assaults/"}},"_tags":["comment","author_thelastgallon","story_49951881"],"author":"thelastgallon","children":[49958896,49954007,49952505,49952618,49953914,49953500,49952282],"comment_text":"Fossil fuel supply chain makes your country very vulnerable!
https://github.com/Lkruitwagen/global-fossil-fuel-supply-cha...: The resulting complex network has 6.09mn nodes and 15.70mn edges and is implemented in a graph database. With estimates of annual coal, gas, and oil demand in 13,229 global population centres and 8,165 global power stations, we use a minimum-cost flow method to estimate global asset-level energy flows.
If you want to protect your country, here are your options:
1) Invest in nuclear power and nukes (source of Uranium may be a concern). See North Korea vs Iran.
2) Solar on every structure, every vehicle is an EV. This will be the most resilient, distributed grid that can never be disrupted. Node level failures (hail on roof breaks solar) can be easily overcome. EVs are batteries on wheels! They can go charge and come back. Some on this forum may be familiar with Cattle vs Pet, Redundancy, Resiliency, Virtualization (--> Virtual Power Plants), etc -- same concepts, but applied to power generation which is designed as a monolithic circuit switched network. The idea of energy as packets, which can be Store-and-Forward (using batteries) will change everything we know about energy.","created_at":"2026-10-04T09:36:09Z","created_at_i":1791106569,"objectID":"49952179","parent_id":49951881,"story_id":49951881,"story_title":"In Ukraine, distributed renewables foil Russia's assaults","story_url":"https://energytransition.org/2026/09/in-ukraine-distributed-renewables-foil-russias-assaults/","updated_at":"2026-10-05T09:10:42Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rumatoest"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"Rust engineer with experience in blockchains core, L1/L2 networks, node infrastructure, crypto payment integrations.
Hands-on experience in building a new blockchain network from scratch in Rust with my own smart contract engine.
Outside blockchain worked on production backend systems, billing systems, big data pipelines, cloud infrastructure and other stuff.
-
Location: Europe (outside EU)
Remote: Global remote (the smallest workday overlap is with PST/PDT)
Willing to relocate: NO, can visit EU if needed.
Technologies: Rust, blockchain core, EVM, Solidity, Solana, Nodes infrastructure; Java/Kotlin, TypeScript, Python.
R\u00e9sum\u00e9/CV: Available upon request via email.
Email: br1bcpmbe@mozmail.com
NOTICE: By default, treat any job offer as a SCAM! No offense, these are hard times."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Who wants to be hired? (October 2026)"}},"_tags":["comment","author_rumatoest","story_49922568"],"author":"rumatoest","comment_text":"Rust engineer with experience in blockchains core, L1/L2 networks, node infrastructure, crypto payment integrations.
Hands-on experience in building a new blockchain network from scratch in Rust with my own smart contract engine.
Outside blockchain worked on production backend systems, billing systems, big data pipelines, cloud infrastructure and other stuff.
-
Location: Europe (outside EU)
Remote: Global remote (the smallest workday overlap is with PST/PDT)
Willing to relocate: NO, can visit EU if needed.
Technologies: Rust, blockchain core, EVM, Solidity, Solana, Nodes infrastructure; Java/Kotlin, TypeScript, Python.
R\u00e9sum\u00e9/CV: Available upon request via email.
Email: br1bcpmbe@mozmail.com
NOTICE: By default, treat any job offer as a SCAM! No offense, these are hard times.","created_at":"2026-10-04T07:57:21Z","created_at_i":1791100641,"objectID":"49951657","parent_id":49922568,"story_id":49922568,"story_title":"Ask HN: Who wants to be hired? (October 2026)","updated_at":"2026-10-04T08:26:39Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jdw64"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"I think it largely depends on individual workflows, and there is no single right answer. Software is at once deeply personal, inherently collaborative, and heavily shaped by communication dynamics.
Naur argued that when a team disbands, the program dies. But in an era where the primary maintainer is no longer human, but an AI agent, how should we approach this?
My take is that we will lean heavily on concepts like Algebraic Data Types (ADTs) and state machines\u2014making state definitions something humans can readily track through explicit specifications\u2014while modularizing boundaries so that humans only need to retain the interfaces.
In other words, rather than keeping low-level implementation details in our heads, systems will be structured around high-level interfaces, data flows, core capabilities, and first-class functions.
Of course, the prevailing assumption today is still that humans must grasp the underlying implementation to work effectively. There are trade-offs to both models, and the optimal balance will naturally vary across different workflows.
Ultimately, the core issue is this: when something goes wrong within the AI's internal implementation, will we actually be able to steer it? And as you know, even ADTs must eventually change when business requirements shift. I haven't thought deeply about how to handle that yet, either.
It's an open question that neither experienced programmers like yourself nor less experienced ones like me have the definitive answer to. Software is heavily dependent on highly personal contexts. My answers and experiences won't neatly apply to your context, and vice versa.
My guess is that we might eventually need an entirely new programming language. What would that language look like? I have no idea. IDEs might also shift away from traditional text blocks toward a UI-based, flowchart-like approach. I imagine a system somewhat like n8n, where we can visualize and track the workflows and code blocks generated by the AI.
I recently did a few "vibe coding" projects with AI. It churned out 50,000 to 60,000 lines of code in an instant, and I couldn't keep up. To cope, I built a tool to visualize and track the code's dependencies. It worked out much better than I expected.
The caveat, of course, is that with tools like Unreal Blueprints or n8n, the visual dependency graph eventually explodes as the program grows, turning the UI itself into a bottleneck. Perhaps the solution will be that programs up to a certain size are abstracted into single nodes, which can then be clicked to expand and reveal the underlying implementation details."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Agents don't need memory, they need documentation"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://liao.gg/blog/agents-dont-need-memory"}},"_tags":["comment","author_jdw64","story_49945933"],"author":"jdw64","comment_text":"I think it largely depends on individual workflows, and there is no single right answer. Software is at once deeply personal, inherently collaborative, and heavily shaped by communication dynamics.
Naur argued that when a team disbands, the program dies. But in an era where the primary maintainer is no longer human, but an AI agent, how should we approach this?
My take is that we will lean heavily on concepts like Algebraic Data Types (ADTs) and state machines\u2014making state definitions something humans can readily track through explicit specifications\u2014while modularizing boundaries so that humans only need to retain the interfaces.
In other words, rather than keeping low-level implementation details in our heads, systems will be structured around high-level interfaces, data flows, core capabilities, and first-class functions.
Of course, the prevailing assumption today is still that humans must grasp the underlying implementation to work effectively. There are trade-offs to both models, and the optimal balance will naturally vary across different workflows.
Ultimately, the core issue is this: when something goes wrong within the AI's internal implementation, will we actually be able to steer it? And as you know, even ADTs must eventually change when business requirements shift. I haven't thought deeply about how to handle that yet, either.
It's an open question that neither experienced programmers like yourself nor less experienced ones like me have the definitive answer to. Software is heavily dependent on highly personal contexts. My answers and experiences won't neatly apply to your context, and vice versa.
My guess is that we might eventually need an entirely new programming language. What would that language look like? I have no idea. IDEs might also shift away from traditional text blocks toward a UI-based, flowchart-like approach. I imagine a system somewhat like n8n, where we can visualize and track the workflows and code blocks generated by the AI.
I recently did a few "vibe coding" projects with AI. It churned out 50,000 to 60,000 lines of code in an instant, and I couldn't keep up. To cope, I built a tool to visualize and track the code's dependencies. It worked out much better than I expected.
The caveat, of course, is that with tools like Unreal Blueprints or n8n, the visual dependency graph eventually explodes as the program grows, turning the UI itself into a bottleneck. Perhaps the solution will be that programs up to a certain size are abstracted into single nodes, which can then be clicked to expand and reveal the underlying implementation details.","created_at":"2026-10-04T07:29:16Z","created_at_i":1791098956,"objectID":"49951515","parent_id":49950732,"story_id":49945933,"story_title":"Agents don't need memory, they need documentation","story_url":"https://liao.gg/blog/agents-dont-need-memory","updated_at":"2026-10-05T00:40:26Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"lukewarm707"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"EA is integral and indispensible to the AI safety complex. Almost all nonprofits, research institutes, evaluators and academics in this field are steered by EA ideology and funding. As far fetched as it sounds it is not an exaggeration.
On funding: the three or four core funding nodes linking this together are EA vehicles at two hops or less between each other and every other major node in the ai safety 'complex'. EA funds almost all of it.
On top of that, there are personal EA connections and the revolving door between the ai industry and the nonprofits. Here are some examples:
Government advisors and regulators. NIST CAISI is the USA Government advisory body. Christiano was head of safety and advises. He is ex-OpenAI, former Amodei associate. His vehicle ARC was on the Coefficient EA payroll. Barnes and Christiano's vehicle Arc Evals similarly received EA cash out of Coefficient, rolling this into what is now METR. Christiano's spouse Cotra worked at Coeffiecient steering EA funding to organizations such as METR, then rotated through the revolving door onto the payroll at METR itself, where she co-authored the oai-hf report.
Many UK AISI advisors are Anthropic and EA associates. Chair Hogarth cashed out of Anthropic. Shlegeris of Redwood Research is an advisor, ex-MIRI (Yudkowsky vehicle). Redwood is funded by the exact same funding triangle: Coefficient, Taallin, FTX/Alameda. Alameda CEO Caroline Ellison dated Shlegeris, then dated FTX CEO Sam Bankman-Fried, then rotated through the revolving door out of prison into formerly FTX-funded Manifund. All EA. AI safety charities were on island retreat in the Bahamas with FTX. Why does AI safety charity Lighthouse own $20m of SF real estate?
Redwood Chief Scientist Ryan Greenblatt (Coefficient funded) co-wrote the oai report with METR; he is married to METR founder Beth Barnes (Coefficient funded).
Coefficient was run by long-time Amodei associate Karnofsky. Karnofsky lived with the Amodeis and is married to Anthropic Board member Daniella Amodei. Karnofsky is now directly on the Anthropic payroll; Coefficient is propped up by Anthropic share value.
Everyone here has been funded one step away from Anthropic cash; they are now proposing to integrate themselves in the government (NIST) and evaluate Anthropic (METR and Redwood).
It is hard to find academics here who have not been deeply embedded in funded EA institutes or Toby Ord vehicles; yet harder to find academics here NOT taking EA grant money. the safety doomer kingpins: Kokotajlo has a executive position at AI Futures, Taallin funded. Benigo has scientific director of LawZero, same series A Anthropic funders who are sitting on a 1000x return (Tallinn, Moskovitz, Schmidt).
These connections and funding are at one or two hops, they are often direct connections. You are looking at a massive swamp network that is really impossible to parse without a lot of work."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"LeCun has \"zero concerns\" about AI wiping out humanity, recent \"rogue\" incidents"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://fortune.com/2026/10/01/ai-godfather-yann-lecun-has-zero-concerns-about-human-extinction-says-anthropic-ceo-dario-amodei-is-deuded/"}},"_tags":["comment","author_lukewarm707","story_49946228"],"author":"lukewarm707","comment_text":"EA is integral and indispensible to the AI safety complex. Almost all nonprofits, research institutes, evaluators and academics in this field are steered by EA ideology and funding. As far fetched as it sounds it is not an exaggeration.
On funding: the three or four core funding nodes linking this together are EA vehicles at two hops or less between each other and every other major node in the ai safety 'complex'. EA funds almost all of it.
On top of that, there are personal EA connections and the revolving door between the ai industry and the nonprofits. Here are some examples:
Government advisors and regulators. NIST CAISI is the USA Government advisory body. Christiano was head of safety and advises. He is ex-OpenAI, former Amodei associate. His vehicle ARC was on the Coefficient EA payroll. Barnes and Christiano's vehicle Arc Evals similarly received EA cash out of Coefficient, rolling this into what is now METR. Christiano's spouse Cotra worked at Coeffiecient steering EA funding to organizations such as METR, then rotated through the revolving door onto the payroll at METR itself, where she co-authored the oai-hf report.
Many UK AISI advisors are Anthropic and EA associates. Chair Hogarth cashed out of Anthropic. Shlegeris of Redwood Research is an advisor, ex-MIRI (Yudkowsky vehicle). Redwood is funded by the exact same funding triangle: Coefficient, Taallin, FTX/Alameda. Alameda CEO Caroline Ellison dated Shlegeris, then dated FTX CEO Sam Bankman-Fried, then rotated through the revolving door out of prison into formerly FTX-funded Manifund. All EA. AI safety charities were on island retreat in the Bahamas with FTX. Why does AI safety charity Lighthouse own $20m of SF real estate?
Redwood Chief Scientist Ryan Greenblatt (Coefficient funded) co-wrote the oai report with METR; he is married to METR founder Beth Barnes (Coefficient funded).
Coefficient was run by long-time Amodei associate Karnofsky. Karnofsky lived with the Amodeis and is married to Anthropic Board member Daniella Amodei. Karnofsky is now directly on the Anthropic payroll; Coefficient is propped up by Anthropic share value.
Everyone here has been funded one step away from Anthropic cash; they are now proposing to integrate themselves in the government (NIST) and evaluate Anthropic (METR and Redwood).
It is hard to find academics here who have not been deeply embedded in funded EA institutes or Toby Ord vehicles; yet harder to find academics here NOT taking EA grant money. the safety doomer kingpins: Kokotajlo has a executive position at AI Futures, Taallin funded. Benigo has scientific director of LawZero, same series A Anthropic funders who are sitting on a 1000x return (Tallinn, Moskovitz, Schmidt).
These connections and funding are at one or two hops, they are often direct connections. You are looking at a massive swamp network that is really impossible to parse without a lot of work.","created_at":"2026-10-04T03:41:55Z","created_at_i":1791085315,"objectID":"49950383","parent_id":49949272,"story_id":49946228,"story_title":"LeCun has \"zero concerns\" about AI wiping out humanity, recent \"rogue\" incidents","story_url":"https://fortune.com/2026/10/01/ai-godfather-yann-lecun-has-zero-concerns-about-human-extinction-says-anthropic-ceo-dario-amodei-is-deuded/","updated_at":"2026-10-04T21:38:41Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tebrun"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"Opus 5.5 is great and cheaper if you compare to fable with close quality in coding (tested in refactoring java to nodejs), but I do not understand why the week before the release Opus 5 started hallucinating (long loop and waste of token for single tasks)"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Getting the most out of Opus 5.5 in Claude and Claude Code"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://claude.dev/blog/getting-the-most-out-of-opus-5-5/"}},"_tags":["comment","author_tebrun","story_49946567"],"author":"tebrun","comment_text":"Opus 5.5 is great and cheaper if you compare to fable with close quality in coding (tested in refactoring java to nodejs), but I do not understand why the week before the release Opus 5 started hallucinating (long loop and waste of token for single tasks)","created_at":"2026-10-03T21:19:30Z","created_at_i":1791062370,"objectID":"49947891","parent_id":49946567,"story_id":49946567,"story_title":"Getting the most out of Opus 5.5 in Claude and Claude Code","story_url":"https://claude.dev/blog/getting-the-most-out-of-opus-5-5/","updated_at":"2026-10-03T23:31:38Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"user43928"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"Using AI for the ComfyUI API or workflows worked pretty well for me even early this year.
However, today I don't see a reason to use ComfyUI at all.
For Qwen Image 2.1, I had Opus 5.5 create a backend outside of ComfyUI and it was able to make generation take 20% less time with some optimizations.
The optimizations it implemented were caching the text computation in Qwen Image 2.1 rather than including it in every step, fusing projections into a larger matrix multiplication, and decoding the VAE in horizontal bands or something like that.
If there was anything interesting in ComfyUI nodes, I imagine I could just have the AI adopt the relevant code instead of dealing with ComfyUI or custom nodes."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"FLUX 3 Image"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://bfl.ai/models/flux-3-image"}},"_tags":["comment","author_user43928","story_49925974"],"author":"user43928","comment_text":"Using AI for the ComfyUI API or workflows worked pretty well for me even early this year.
However, today I don't see a reason to use ComfyUI at all.
For Qwen Image 2.1, I had Opus 5.5 create a backend outside of ComfyUI and it was able to make generation take 20% less time with some optimizations.
The optimizations it implemented were caching the text computation in Qwen Image 2.1 rather than including it in every step, fusing projections into a larger matrix multiplication, and decoding the VAE in horizontal bands or something like that.
If there was anything interesting in ComfyUI nodes, I imagine I could just have the AI adopt the relevant code instead of dealing with ComfyUI or custom nodes.","created_at":"2026-10-03T11:58:38Z","created_at_i":1791028718,"objectID":"49943490","parent_id":49940655,"story_id":49925974,"story_title":"FLUX 3 Image","story_url":"https://bfl.ai/models/flux-3-image","updated_at":"2026-10-03T13:26:51Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"nodesocket"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"Fair, I get the performance of DX12 or OpenGL may vary widely, but I was thinking for AI applications though further research on my part showed a dramatic math difference.
Sha256 is mostly integer additions, bit rotations, XOR, AND operations. While AI workload are matrix multiplication and accumulation. I sort of answered my own question then. :)"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Why don't we just use hashcat to evaluate GPU performance"}},"_tags":["comment","author_nodesocket","story_49942621"],"author":"nodesocket","comment_text":"Fair, I get the performance of DX12 or OpenGL may vary widely, but I was thinking for AI applications though further research on my part showed a dramatic math difference.
Sha256 is mostly integer additions, bit rotations, XOR, AND operations. While AI workload are matrix multiplication and accumulation. I sort of answered my own question then. :)","created_at":"2026-10-03T09:34:01Z","created_at_i":1791020041,"objectID":"49942695","parent_id":49942630,"story_id":49942621,"story_title":"Ask HN: Why don't we just use hashcat to evaluate GPU performance","updated_at":"2026-10-03T09:35:06Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nodejs"],"value":"nodesocket"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"I ran the following on a M4 Pro (20 core GPU) in MacOS vs a 4070 Super in Windows 11 using WSL and was kinda shocked to see the dramatic difference in GPU performance.
## generate a random sha256 8 alphanumeric\n python3 - <<'PY' > hashes-sha256.txt\n import secrets\n import string\n import hashlib\n\n alphabet = string.ascii_letters + string.digits\n password = ''.join(secrets.choice(alphabet) for _ in range(8))\n print(hashlib.sha256(password.encode()).hexdigest())\n hashcat -m 1400 -a 3 -O \\\n --potfile-path eight-alphanumeric-sha256.pot \\\n -1 '?l?u?d' \\\n hashes-sha256.txt '?1?1?1?1?1?1?1?1'\n ## generate a random sha256 8 alphanumeric\n python3 - <<'PY' > hashes-sha256.txt\n import secrets\n import string\n import hashlib\n\n alphabet = string.ascii_letters + string.digits\n password = ''.join(secrets.choice(alphabet) for _ in range(8))\n print(hashlib.sha256(password.encode()).hexdigest())\n hashcat -m 1400 -a 3 -O \\\n --potfile-path eight-alphanumeric-sha256.pot \\\n -1 '?l?u?d' \\\n hashes-sha256.txt '?1?1?1?1?1?1?1?1'\n