Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Dream-RSI: Recursive Self-Improvement through Evolving Worlds(arxiv.org ↗)
  2. Show HN: Give your AI agents access to WhatsApp
  3. Mistral X Mozilla: Private, Multilingual AI Browsing(mistral.ai ↗)
  4. The Google Play app review process now regularly takes longer than a week(gultsch.social ↗)
  5. Introducing System One Models and Jev(typesafe.ai ↗)
  6. Show HN: An e-ink frame that hears birds and draws them as 1800s illustrations(github.com/arnegiacomo ↗)
  7. Hackers Got Inside a Flock Camera. Its Data Shows How the System Works(wired.com ↗)
  8. How Big Are Factorials?(thegreenplace.net ↗)
  9. Apple Reference Image: A New Approach for Verified Photography(security.apple.com ↗)
  10. Scaling Golang CI by Replacing actions/setup-go(cloudx.ai ↗)
  11. Douglas Adams and the exterminated Doctor Who adventure(bbc.co.uk ↗)
  12. Original Sony PlayStation 2 security chip 'broken wide open' after 26 years(tomshardware.com ↗)
  13. PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand"(frvr.com ↗)
  14. Kyber (YC W23) Is Hiring a Forward Deployed Engineer(ycombinator.com ↗)
  15. Tell the speakers that you liked their talks(ohhelloana.blog ↗)
  16. An update on Wayback Machine access(blog.archive.org ↗)
  17. Microsoft says AI rival Anthropic could have 'disastrous impact' on humanity(bbc.co.uk ↗)
  18. Salesforce Global Outage(salesforce.com ↗)
  19. Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models(stale.jock.pl ↗)
  20. ImpactGate: A merge gate that scores the structural decay AI adds(github.com/officefloor ↗)
  21. Doing Everyone Else's Job(yosefk.com ↗)
  22. DeepSeek v4.1 Flash Is Now Our Best Hacking Model(enclave.ai ↗)
  23. Show HN: I made a flight simulator, except you're just a passenger(inflightsimulator.com ↗)
  24. OpenAI expands ChatGPT ads with Sponsored Agents(openai.com ↗)
  25. Gemini 3.8 Live and 3.8 Live Extended Thinking(blog.google ↗)
  26. Why I'm still bearish on LLMs after Navier-Stokes(dank.systems ↗)
  27. Intelligence per Watt: Measuring Intelligence Efficiency of Local AI(arxiv.org ↗)
  28. Anatomy of a Texture(agentlien.github.io ↗)
  29. A software thing I built: GPS on a 25MHz 486-SX(vcfed.org ↗)
  30. German Rheinmetall open-sources its Battlesuite connected weapon system protcol(rheinmetall.github.io ↗)

Dream-RSI: Recursive Self-Improvement through Evolving Worlds

55 pointsby 2h agoarxiv.org
15 comments
2h agoHN ↗

Would be very glad if anyone explained to me if and why this is revolutionary.

1h agoHN ↗

Here you go:

"Across algorithm engineering, mathematical optimization, and GPU kernel engineering, Dream-RSI achieves competitive or improved discovery quality while substantially reducing discovery cost in several settings."

1h agoHN ↗

Probably isn't by virtue of it being publicly released

1h agoHN ↗

Fairly certain all the labs are doing this (RSI) at this point. It's a question of how public their proclamations are about it and how they're positioning PR etc.

Even today's lighter weight models know how to write kernels and optimize them. I've had DeepSeek 4.1 Flash tune the crap out custom CUDA kernels on my own codebase and it was entirely competent at it. And cheap.

The innovation pieces will be in the harnesses to support this. Which I guess is partially what's going on here.

1h agoHN ↗

It's not really RSI if you are just using the AI as a tool to help make it better. It has to be doing it itself, no? Otherwise self-hosted compilers are RSI.

1h agoHN ↗

Why would you assume it is revolutionary?

21m agoHN ↗

Because “recursive self-improvement” is widely understood, even at labs, as having a non-trivial chance of eradicating human life.

This is like asking why a paper called “I discovered God. This is the specific proof.” is revolutionary. Just tedious levels of incredulity.

1h agoHN ↗

Unless I'm misunderstanding, calling this RSI seems misleading?

This looks like an optimization of current training methods, and a good one, but not "RSI" in the sense of a system that can perpetually improve itself forever.

20m agoHN ↗

Does RSI actually mean anything specific anymore? RSI, AGI, at this point seem like buzzwords. Sure AGI has definition that are measurable, say "better than 95% of humans on 95% of intellectual tasks" but if we used that definition we already have AGI and almost no one thinks we have achieved AGI. We use AIs to train AIs which we use to train AIs, why is that not RSI? How much human intervention means that is not RSI?

14m agoHN ↗

Yeah, this is absolutely not what anyone reasonable is thinking about when they say recursive self-improvement.

I'd say it's much closer to the concept of continual learning, but I'm only a few pages deep and haven't groqued it fully yet.

49m agoHN ↗

This is a solid and very interesting paper! The authors were kind enough to publish the complete prompt for it (appendix B.1 on page 18), so anyone can try their approach with any LLM and see the results.

26m agoHN ↗

So its more like a harness (like Codex etc) on top of any existing LLM?

17m agoHN ↗

I ended up building a simplified version of this as /self-improve in https://github.com/DanMcInerney/orchflows. History is the state ledger, memory and RSI just cite the history as evidence and can be rewritten. I feel like strong immutable state is the missing piece of the puzzle for most of these memory libraries.

13m agoHN ↗

I might be wrong but is this really a solution for RSI? I interpret it's as a way to reducing wasted tokens and compute on paths that don't yield better results. It's an optimization. It's a faster way to get to RSI though. what's wrong?