Toby Ord paper argues recursive self-improvement has physical limits
Exponential View (Azeem Azhar) · rss · 2026-08-30
Toby Ord's new paper explores the possibility and limits of Recursive Self-Improvement (RSI). Ord argues that while unbounded intelligence growth is mathematically possible, it is practically unlikely due to the 'generation time' constraint—the time needed for an AI to complete an improvement loop (designing and training its successor). Real-world constraints like experiments, training runs, and chip manufacturing prevent generation time from reaching zero. Furthermore, physical laws such as the speed of light, the Bekenstein bound, and Landauer's principle impose ultimate limits on computation speed and information density, making unbounded RSI improbable.
More from AGI Musings
- First-ever paper on multi-agent swarm safety lands NeurIPS spotlight — LauraRuis · 2026-10-16
- Yoav Artzi on peer review: paper flooding still pays off as review system strains under load — yoavartzi · 2026-10-16
- Letta founder: work in fields defined by problems, not methods — lateinteraction · 2026-10-16
- Why pure mathematicians will soon flood into AI x-risk debates — Dr_Atoosa · 2026-10-16
- RSI began when model outputs augmented next-gen training data, agents just escalated it — akbirthko · 2026-10-16
- Luiza Jarovsky: with AI making careers unpredictable, 'do what you love' is the best advice left — LuizaJarovsky · 2026-10-16