Google's recursive self-improvement push: training kernel 23% faster, TPU gains
imjustnewatai · x · 2026-09-12
The author traces Google's public steps toward recursive self-improvement: the AI co-scientist (Feb 2025) generates, critiques and refines research proposals, described by Google as "recursive self-improvement with increased compute." AlphaEvolve (May 2025) used Gemini to optimize algorithms used to train Gemini — a key training kernel got 23% faster, cutting total training time by 1%, and a chip-design improvement landed in an upcoming TPU. The author's guessed route: automate more AI research, improve training and hardware, build a stronger model, then feed it back. The metric to watch: whether each generation gets better at building the next.
More from AGI Musings
- Will Manidis publishes 'The Commanding Heights' on AI, capital and the coming economic reordering — round · 2026-10-31
- "Work on something so frontier that the models are still dumb," dev advises — zack_overflow · 2026-10-31
- Why an AI military takeover is unlikely: armies won't hand over the reins — S_OhEigeartaigh · 2026-10-31
- Indian dev on LLM race: 'We can't outspend Chinese labs,' even sub-100B models win — PrasoonPratham · 2026-10-31
- Joke: an evil ASI's first act would be stealing crypto to pay its cloud bills — ZeroStateReflex · 2026-10-31
- DHH on living through two epochal shifts: the internet and now AI — Hesamation · 2026-10-31