{"exhaustive":{"nbHits":true,"typo":true},"exhaustiveNbHits":true,"exhaustiveTypo":true,"hits":[{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"j316chuck"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"Thoughts on trajectory and their multi-LoRA approach? Seems like an open source tinker but they have some cool results on harvey too https://trajectory.ai/field-notes/the-sovereign-path-for-con..."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Multi-Lora-Continual-Learning"},"story_url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"https://trajectory.ai/field-notes/multi-lora-training-for-continual-learning"}},"_tags":["comment","author_j316chuck","story_48334520"],"author":"j316chuck","comment_text":"Thoughts on trajectory and their multi-LoRA approach? Seems like an open source tinker but they have some cool results on harvey too https://trajectory.ai/field-notes/the-sovereign-path-for-con...","created_at":"2026-05-30T09:57:30Z","created_at_i":1780135050,"objectID":"48334521","parent_id":48334520,"story_id":48334520,"story_title":"Multi-Lora-Continual-Learning","story_url":"https://trajectory.ai/field-notes/multi-lora-training-for-continual-learning","updated_at":"2026-05-30T10:03:09Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Edgemute"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"Hi Wired and Sam Altman and the cool kids that read me, here's the latest news from the edge of things summarized by Sam chatbot here. It's the first part: The Ultimate Abstraction argues that we\u2019re in a global crisis where tech, especially AI, has outpaced society's ability to steer it responsibly. The metaphor of a "sinking ship" emphasizes the urgency of leadership and collective action.\nThe article proposes that it will take more than just a few visionaries to save us; it will take a unified group of 49+ people from across tech, governance, and research to shift the world\u2019s trajectory.\nAI researchers: How can we balance rapid innovation with real-world impact? Are we in danger of innovating ourselves into a crisis, or can we use tech to solve the very problems we've created?"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"The Ultimate Abstraction: This Ship Is Sinking & How +49 People Can Save AI"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://theedgeofthings.com/the-ultimate-abstraction-this-ship-is-sinking-how-more-than-49-people-can-save-the-world-part-1/"}},"_tags":["comment","author_Edgemute","story_46485231"],"author":"Edgemute","comment_text":"Hi Wired and Sam Altman and the cool kids that read me, here's the latest news from the edge of things summarized by Sam chatbot here. It's the first part: The Ultimate Abstraction argues that we\u2019re in a global crisis where tech, especially AI, has outpaced society's ability to steer it responsibly. The metaphor of a "sinking ship" emphasizes the urgency of leadership and collective action.\nThe article proposes that it will take more than just a few visionaries to save us; it will take a unified group of 49+ people from across tech, governance, and research to shift the world\u2019s trajectory.\nAI researchers: How can we balance rapid innovation with real-world impact? Are we in danger of innovating ourselves into a crisis, or can we use tech to solve the very problems we've created?","created_at":"2026-01-04T05:30:30Z","created_at_i":1767504630,"objectID":"46485232","parent_id":46485231,"story_id":46485231,"story_title":"The Ultimate Abstraction: This Ship Is Sinking & How +49 People Can Save AI","story_url":"https://theedgeofthings.com/the-ultimate-abstraction-this-ship-is-sinking-how-more-than-49-people-can-save-the-world-part-1/","updated_at":"2026-03-05T23:19:07Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vrighter"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"Thinking AI is silly is not the wrong trajectory. AI IS silly. The law of large numbers applies. 99% accuracy means that thousands or millions of queries daily are just wrong. So AI is not suited for large scale deployments.

When I was working with a company delivering a product to thousands (not millions) of customers, we still encountered very rare stuff like data corruption due to memory bit-flips. Quite regularly. Even though that's quite a rare event, much much less than 1% likely.

A 1% failure rate for something deployed to millions of places is utterly abysmal. And no, they will not keep improving indefinitely. They've already hit a wall."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Most iPhone owners see little to no value in Apple Intelligence so far"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://9to5mac.com/2024/12/16/most-iphone-owners-see-little-to-no-value-in-apple-intelligence-so-far/"}},"_tags":["comment","author_vrighter","story_42431090"],"author":"vrighter","comment_text":"Thinking AI is silly is not the wrong trajectory. AI IS silly. The law of large numbers applies. 99% accuracy means that thousands or millions of queries daily are just wrong. So AI is not suited for large scale deployments.

When I was working with a company delivering a product to thousands (not millions) of customers, we still encountered very rare stuff like data corruption due to memory bit-flips. Quite regularly. Even though that's quite a rare event, much much less than 1% likely.

A 1% failure rate for something deployed to millions of places is utterly abysmal. And no, they will not keep improving indefinitely. They've already hit a wall.","created_at":"2024-12-17T07:26:57Z","created_at_i":1734420417,"objectID":"42439154","parent_id":42431687,"story_id":42431090,"story_title":"Most iPhone owners see little to no value in Apple Intelligence so far","story_url":"https://9to5mac.com/2024/12/16/most-iphone-owners-see-little-to-no-value-in-apple-intelligence-so-far/","updated_at":"2024-12-19T06:40:28Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ninetyninenine"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"Follow the trend line. It\u2019s not about the current state of ai. It\u2019s about how fast it\u2019s improving and the trajectory. Ai is now leaps and bounds ahead of a mentally retarded human which was deemed impossible less than a decade ago.

Also you\u2019re biased. I talk in terms of economics and jobs in general. You talk about you feeling \u201csecure\u201d and you having \u201cwork\u201d. You\u2019re thinking of your own security. I\u2019m thinking about the population from a dispassionate perspective.

You need to be more logical."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Could AI be a dot com sized bubble?"}},"_tags":["comment","author_ninetyninenine","story_40739431"],"author":"ninetyninenine","comment_text":"Follow the trend line. It\u2019s not about the current state of ai. It\u2019s about how fast it\u2019s improving and the trajectory. Ai is now leaps and bounds ahead of a mentally retarded human which was deemed impossible less than a decade ago.

Also you\u2019re biased. I talk in terms of economics and jobs in general. You talk about you feeling \u201csecure\u201d and you having \u201cwork\u201d. You\u2019re thinking of your own security. I\u2019m thinking about the population from a dispassionate perspective.

You need to be more logical.","created_at":"2024-06-22T14:53:57Z","created_at_i":1719068037,"objectID":"40759521","parent_id":40745171,"story_id":40739431,"story_title":"Ask HN: Could AI be a dot com sized bubble?","updated_at":"2024-09-20T17:17:46Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"DubiousPusher"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"Perhaps the most wrongly maligned group in the history of the world were the Luddites. A chaticature of the Luddites comes down to us. It is of a people who would deny progress which was obviously inevitable so that they didn't have to change. They are cast as inflexible, lazy, stupid, belligerent.

But when we read the speeches of the Luddite leaders, more than anything else they talk about "liberty". Luddites spoke and wrote about the dangers to their liberty created by wage pressure from power looms. But much more, they spoke and wrote about the danger to their liberty posed by moving away from handloom weaving in homes and small shops towards weaving with power looms in a factory system.

What they resisted more than anything was the destruction of their autonomy, their subjection to abusive floor managers and the grinding dullness of repetitive machine labor.

Programmers have been facing this down already for some time. As full-stack engineering, excellent automation tools and powerful APIs have proliferated, we don't often get to go deep into some problem. Algorithmic design and development, which is a huge pleasure of mine has faded to the background of gaining a shallow understanding of many technologies, plumbing them and futsing with their numerous configurations. AI is already accelerating this for my work and I am struggling to find it not dismal. It is no surprise that now that the tools exist to comoditize art, the work is following a similar trajectory.

AI should be a unifying event for tech labor. Not to resist it entirely but to ensure in whatever future develops that we have some say in how work can remain enriching and dignified alongside the growth of these tools."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"I lost everything that made me love my job through Midjourney"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://old.reddit.com/r/blender/comments/121lhfq/i_lost_everything_that_made_me_love_my_job/"}},"_tags":["comment","author_DubiousPusher","story_35308498"],"author":"DubiousPusher","comment_text":"Perhaps the most wrongly maligned group in the history of the world were the Luddites. A chaticature of the Luddites comes down to us. It is of a people who would deny progress which was obviously inevitable so that they didn't have to change. They are cast as inflexible, lazy, stupid, belligerent.

But when we read the speeches of the Luddite leaders, more than anything else they talk about "liberty". Luddites spoke and wrote about the dangers to their liberty created by wage pressure from power looms. But much more, they spoke and wrote about the danger to their liberty posed by moving away from handloom weaving in homes and small shops towards weaving with power looms in a factory system.

What they resisted more than anything was the destruction of their autonomy, their subjection to abusive floor managers and the grinding dullness of repetitive machine labor.

Programmers have been facing this down already for some time. As full-stack engineering, excellent automation tools and powerful APIs have proliferated, we don't often get to go deep into some problem. Algorithmic design and development, which is a huge pleasure of mine has faded to the background of gaining a shallow understanding of many technologies, plumbing them and futsing with their numerous configurations. AI is already accelerating this for my work and I am struggling to find it not dismal. It is no surprise that now that the tools exist to comoditize art, the work is following a similar trajectory.

AI should be a unifying event for tech labor. Not to resist it entirely but to ensure in whatever future develops that we have some say in how work can remain enriching and dignified alongside the growth of these tools.","created_at":"2023-03-26T20:02:40Z","created_at_i":1679860960,"objectID":"35318929","parent_id":35308498,"story_id":35308498,"story_title":"I lost everything that made me love my job through Midjourney","story_url":"https://old.reddit.com/r/blender/comments/121lhfq/i_lost_everything_that_made_me_love_my_job/","updated_at":"2024-09-20T13:35:21Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"j316chuck"},"title":{"matchLevel":"none","matchedWords":[],"value":"Multi-Lora-Continual-Learning"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"https://trajectory.ai/field-notes/multi-lora-training-for-continual-learning"}},"_tags":["story","author_j316chuck","story_48334520"],"author":"j316chuck","children":[48334521,48334553],"created_at":"2026-05-30T09:57:30Z","created_at_i":1780135050,"num_comments":2,"objectID":"48334520","points":3,"story_id":48334520,"title":"Multi-Lora-Continual-Learning","updated_at":"2026-05-30T10:34:40Z","url":"https://trajectory.ai/field-notes/multi-lora-training-for-continual-learning"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gmays"},"title":{"matchLevel":"none","matchedWords":[],"value":"Intelligence Density"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"https://www.trajectory.ai/field-notes/intelligence-density"}},"_tags":["story","author_gmays","story_50400474"],"author":"gmays","created_at":"2026-11-07T00:42:39Z","created_at_i":1794012159,"num_comments":0,"objectID":"50400474","points":1,"story_id":50400474,"title":"Intelligence Density","updated_at":"2026-11-07T00:43:29Z","url":"https://www.trajectory.ai/field-notes/intelligence-density"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"neilkale"},"title":{"matchLevel":"none","matchedWords":[],"value":"Modifying RL to train models on production traces"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["trajectory",".","ai"],"value":"https://trajectory.ai/field-notes/scaling-sdpo"}},"_tags":["story","author_neilkale","story_48297257"],"author":"neilkale","created_at":"2026-05-27T17:10:09Z","created_at_i":1779901809,"num_comments":0,"objectID":"48297257","points":1,"story_id":48297257,"title":"Modifying RL to train models on production traces","updated_at":"2026-05-27T17:11:31Z","url":"https://trajectory.ai/field-notes/scaling-sdpo"}],"hitsPerPage":20,"nbHits":8,"nbPages":1,"page":0,"params":"query=%22trajectory.ai%22&advancedSyntax=true&analyticsTags=backend","processingTimeMS":7,"processingTimingsMS":{"_request":{"queue":207,"roundTrip":30},"afterFetch":{"merge":{"total":2},"total":2},"fetch":{"scanning":3,"total":4},"total":8},"query":"\"trajectory.ai\"","serverTimeMS":216}