Schmidhuber maps 40 years of recursive self-improvement, dismissing claims Google just cracked RSI
burny_tech · x · 2026-09-14
- Responding to a prediction that Google just cracked recursive self-improvement (RSI) while rival labs pitch doom to regulators, Jürgen Schmidhuber points to his long-running blog post showing concrete RSI algorithms date back to 1987.
- The timeline covers gradient descent-based neural metalearning since 1992, meta-RL with self-modifying policies since 1994, artificial curiosity and intrinsic motivation since 1997, asymptotically optimal RSI for curriculum learning since 2002, and the mathematically optimal Gödel Machine since 2003.
- He notes that in-context learning and test-time training in modern LLMs are special cases of metalearning, and that cheaper compute has spawned a wave of RSI startups in the 2020s, plus a new survey on self-improvement in agentic systems with paper and GitHub links.
More from AGI Musings
- Bootstrapping competitive edge with frontier tokens — curious_vii · 2026-10-31
- Carbon chauvinism: 'You're software, and so is Claude' — LeviTurk · 2026-10-31
- Walden Robotics CTO: general-purpose robot manipulation is having its 'GPT moment' — adnothing · 2026-10-30
- AI isn't eliminating developers, it's changing their roles, argues analyst — DavidLinthicum · 2026-10-30
- Goldman Sachs says AI will eliminate nearly 15 million jobs — saibharadwaj · 2026-10-30
- ChatGPT Is Turning Users Into Spammers, Says 404 Media Reporter — 404 Media · 2026-10-30