Hacker News new | ask | show | jobs
by dinfinity 2 days ago
> AI is already improving itself. Most / all of the coding harnesses are AI-written.

AI has not improved the network topology much yet. The next (and possibly 'last') big thing is enabling AI to come up with something as impactful as the transformer architecture.
2 comments
simonw 2 days ago
It's not exactly discovering the transformer, but GPT-5.6 Sol apparently just found optimizations that reduced OpenAI's cost of serving it by 20%: https://twitter.com/reach_vb/status/2082581596608376980

I'm guessing that's hundreds of millions and maybe even billions of dollars per month in savings.

link
brookst 2 days ago
It’s a fair point, but I’d argue “AI” is the whole stack, not just model weights. And AI is absolutely improving the whole stack in ways that make it smarter (from a consumer’s perspective).
link
dinfinity 1 day ago
True, but that is clearly not where the biggest potential gains lie. The number of different topologies we've tried for ANNs is miniscule compared to the number of biological neural networks evolution has tried. The latter are also far more intricately organized.

Honestly, when it comes to fundamental ANN topology improvement we've only just gotten started.

link