MIRI endorses Ban Artificial Superintelligence Act, backs training pauses on precursor capabilities
ESYudkowsky · x · 2026-09-24
MIRI published its official position endorsing the Ban Artificial Superintelligence Act of 2026, calling it the first legislation with a real chance of stopping the ASI extinction threat.
Key provisions MIRI likes: a mandate to pause training if models show precursor capabilities like automating AI R&D or evading shutdown, and a compute threshold above which development requires a charter. The group says the bill is excellent but not perfect and outlines tweaks, arguing intervention must happen before superintelligence exists.
More from AGI Musings
- OpenAI's Boaz Barak: no benevolent AI dictators, even if it means less abundance — tszzl · 2026-09-24
- AI speeds up coding inputs — but has product quality actually improved? An open empirical question — salgar · 2026-09-24
- CNAS report maps the spectrum of AGI futures and how to shape them — paul_scharre · 2026-09-24
- Philosopher Carissa Veliz: AI dominance is not inevitable — the future is produced, not predicted — CarissaVeliz · 2026-09-24
- Karpathy's 11-month tone shift: from coining vibe coding to feeling far behind — IgorCarron · 2026-09-24
- Taleb: AI Is Making High-IQ, Low-Imagination People Obsolete — intellectronica · 2026-09-24