Schmidhuber: I published the first concrete RSI algorithms in 1987, now RSI is driving AI's future
SchmidhuberAI · x · 2026-09-17
Jürgen Schmidhuber revisits the history of Recursive Self-Improvement, claiming he published the first concrete RSI algorithms in 1987 when compute was 100M× more expensive, in a new technical note (IDSIA-9-26). He lists earlier work: self-modifying policies (1994), gradient-based RSI in neural nets (1992), asymptotically optimal RSI for curriculum learning (2002), the self-referential Gödel Machine (2003), and RSI combined with artificial curiosity (1990/1997). His take: software-based RSI is now practical and driving both software and physical AI, but full RSI will require self-improving hardware in the physical world.
Related event: Schmidhuber Claims Priority: First Concrete RSI Algorithm Dates to 1987(10 posts)→
More from AGI Musings
- Avoiding AI brain rot: Miles Cranmer's lecture on learning with AI — _ScottCondron · 2026-11-02
- Vitalik's "post-AI cryptography": pseudonyms are dead, per-action anonymity is in — devanshmehta · 2026-11-02
- Can law be formalized as deterministic rules? A French thread argues LLMs unlock the soft cases — IgorCarron · 2026-11-02
- Nobel economist William Nordhaus dies; essays revisit his growth, tech and climate work — Afinetheorem · 2026-11-02
- dhh: agents' great gift is routing around every gatekeeper in software — Stack Overflow proved it — IgorCarron · 2026-11-02
- Unruly Futures 2026 convenes on post-AGI economics and AI-governed institutions — sarahdrinkwater · 2026-11-02