News for hackers. Stuff that matters.
Re:AI researchers debate how close we are to recursive self-... by Buttons840 on Friday September 11, 2026 @09:33PM (#49667732)
People seem to expect a sudden shift with "self-improvement", but don't AIs already improve themselves via training? What is there to improve?
Re:AI researchers debate how close we are to recursive self-... by pennomi on Friday September 11, 2026 @10:49PM (#49668244)
The human brain operates at better levels of intelligence than the best LLMs, at 20 watts of power. We are a long way to that kind of efficiency, it will probably take both bespoke hardware and algorithmic improvements to catch up to nature.
Re:AI researchers debate how close we are to recursive self-... by realaleris149 on Saturday September 12, 2026 @05:53AM (#49670762)
Or some solar panel arrays closer to the sun and it does not matter?
Re:AI researchers debate how close we are to recursive self-... by gchamonlive on Friday September 11, 2026 @09:54PM (#49667874)
Only if you think in terms of perceived raw intelligence, but self-update is a form of valuable self-improvement that could benefit current models a lot, if they could commit facts from context into their weights cheaply and reliably.
Re:AI researchers debate how close we are to recursive self-... by appplication on Friday September 11, 2026 @10:34PM (#49668138)
I think the idea is fundamentally improved architectures. For example, transformer-based models were an incredible stepwise improvement. Self improvement would be a model discovering a stepwise improvement similar to the transformer. And presumably the improved models from that would be more likely to make further advances still.

Learning from training data is technically self-improvement but not the sort that is typically meant in this context.

Re:AI researchers debate how close we are to recursive self-... by zer00eyz on Friday September 11, 2026 @09:44PM (#49667813)
> AIs already improve themselves via training

Marginally. Model collapse is still a problem. Continuous learning is still a problem.

For AI to make a big leap we need a big break through.