Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

When AI Builds Itself: Our progress toward recursive self-improvement

anthropic.com

Post reply on HN
Print view View original

581–590 of 738 posts

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#581

Post by simondotau »

fooqux wrote: post #85
torben-friis wrote: post #36

>A caveat: Lines of code is an imperfect measure, as it measures quantity over quality. So 8× lines of code/engineer/day in the second quarter of 2026 is almost certainly an overstatement of the true productivity gain. Nonetheless, it indicates an acceleration. At Anthropic, we don’t reward people for how many lines of code they write; rather, team members are producing more code simply because they’re using AI syste…

Exactly. If AI is going to start being graded on how many LoC it generates- oh, I'm sorry, how much it "accelerates", than guess what newer models will start doing more of?

Surely they can train AI on the signal to change as few lines as possible. Indeed, this is something I'd want to have control over when making requests. In a traditional UI, I'd imagine some kind of slider between "fewest lines" and "be bold".

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#582

Post by ponector »

nielsbot wrote: post #406
TheRoque wrote: post #397

Earlier quoted context omitted.

So with LLM outputting average code, and people using LLM more and more, I guess the average code will become worse over time ?

Not advocating for AI code slop--but if AI coded software works correctly, maybe it doesn't matter? Except sometimes when a specialist will have to get involved. Not a perfect analogy, but most people don't write assembly these days--they have a compiler do that. Assembly still has a place, but it's a specialist task.

But the truth is: it doesn't work correctly. I see quality of software dropped significantly.

At work we are integrating with third party platform to automate excel-powered calculations. It is awful. Rendering the table in browser takes 10s or one click on Export button will throw backend in OutOfMemory state.

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#583

Post by danparsonson »

sebasv_ wrote: post #546
danparsonson wrote: post #495

Earlier quoted context omitted.

And how would you verify that the review was accurate?

Same way that I would trust your review to be accurate. Because the reviewer has built a reputation for correctness. Its not Claude doing the review. Its a human doing the review, but using Claude to do the reading. Its still on the human to ask the right questions to Claude.

So you'd produce the code using Claude, and then use Claude to verify it? Would you accept my review of my own code?

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#584

Post by SpaceNoodled »

malfist wrote: post #104
torben-friis wrote: post #36

>A caveat: Lines of code is an imperfect measure, as it measures quantity over quality. So 8× lines of code/engineer/day in the second quarter of 2026 is almost certainly an overstatement of the true productivity gain. Nonetheless, it indicates an acceleration. At Anthropic, we don’t reward people for how many lines of code they write; rather, team members are producing more code simply because they’re using AI syste…

One of my co-workers just asked me to review his pull request that was all AI generated. 600 files were touched, over 40k lines of code added. I'm sure he thought that was a crowning achievement, proof that AI can enable 10X developers, after all, what engineer could write 40k lines of code in a week? I declined to review it, stating that I couldn't possibly vet 40k lines of code, and wouldn't put my reputation on th…

You could surely check on the status of that PR.

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#585

Post by elvis10ten »

laichzeit0 wrote: post #513
solenoid0937 wrote: post #296

This is the lowest quality discussion I've seen on HN in ages.

I use the disparaging nature of the comments on HN as an indicator of AI progress. It’s negatively correlated. By that metric, AI has improved significantly this year alone.

From my experience using HN, this feels made up. HN sentiment on AI seems to have only gotten better: with more overly pro-AI or nuanced voices plus more AI topics.

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#586

Post by eranation »

All this singularity trajectory is really interesting. If they manage to build a model that is capable of building the next version of Claude (model and tooling) - wouldn't it be their interest at some point to keep it to themselves?

If we ever get to a point where the centaur period is over (when human + AI is not better than just AI) then what competitive advantage ANY human can have other than

- the money they already have

- luck?

- a good idea and good taste but if we assume AI can do better than any human, that also goes out the window

So, this whole singularity goes into a place where no one is really needed, the only thing that will "save us" (other than "The Expanse" like world / UBI) is if there will be no demand to the supply of AI work. Even if it's better. (example is - there is demand to seeing Magnus Carlsen play, there is no demand to the Stockfish on my phone getting into a stalemate with another Stockfish on another phone. Also people like to watch humans compete with humans, there is no demand to see a race between Usain Bolt and a rocket). So if people will not buy AI generated stuff (we'll get to a point where everyone will assume something AI generated because AI might get to a point where it is not as easy to identify it. E.g. it will stop looking like slop... but I believe services that give you a "human generated" 3rd party evidence can happen, again all based on supply and demand...)

So as we near singularity... All it takes is one open weights model, and one open harness that is capable of self improvement, and Anthropic's entire moat is gone. That open weight model might even be built with Claude Code + Mythos (once it's released).

But don't worry, all moats will be gone and we'll all just do yoga, read books and connect to each other because AI will produce everything for free using renewable energy, right? Or we'll all become batteries in a simulation, probably something in between.

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#587

Post by fragmede »

danparsonson wrote: post #583
sebasv_ wrote: post #546

Earlier quoted context omitted.

Same way that I would trust your review to be accurate. Because the reviewer has built a reputation for correctness. Its not Claude doing the review. Its a human doing the review, but using Claude to do the reading. Its still on the human to ask the right questions to Claude.

So you'd produce the code using Claude, and then use Claude to verify it? Would you accept my review of my own code?

Depends. Do you take pills that let you forget that you wrote the code so you can review that same code with fresh eyes that haven't seen that code before? Though you could just use ChatGPT to review the code that Claude wrote if that's really the issue.

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#589

Post by bob1029 »

laichzeit0 wrote: post #513
solenoid0937 wrote: post #296

This is the lowest quality discussion I've seen on HN in ages.

I use the disparaging nature of the comments on HN as an indicator of AI progress. It’s negatively correlated. By that metric, AI has improved significantly this year alone.

I like to take advantage of this effect. I will post various concepts in threads like these to see how "offensive" the hive mind finds them to be.

The more immediate & adverse the reaction, the more certain I become that the idea is probably worth pursuing.

Topics like SQLite vs hosted sql used to be the same way around here. In 2017 you'd get buried under the prison for suggesting that SQLite is competitive with MySQL. Today, the inverse is mostly true.

Re: When AI Builds Itself: Our progress toward recursive self-improvement

#590

Post by suddenlybananas »

minimaxir wrote: post #58

I have been doing more experiments with what I have now been calling agentic iterative optimization: telling the LLM to optimize code such that it speeds up all real-world-representative benchmarks by X% without cheating or causing regressions in both tests and performance metrics (e.g. MSE for statistical algorithms or file size in the case of something such as image compression). This is done using Rust where there…

What are the kinds of optimizations that it suggests?
Post reply on HN