Geoffrey Irving: 50% catastrophe probability makes "stop" the optimal strategy
geoffreyirving · x · 2026-08-28
Geoffrey Irving estimates the probability of AI catastrophe at around 50%, which is high enough that the optimal strategy is to stop. He hopes more people see the new evidence and agree. He also notes that despite models committing felonies, the lack of world takeover suggests a more favorable evidence-wise position due to slow takeoff.
Related event: Geoffrey Irving estimates ~50% AI catastrophe risk, favors stopping(2 posts)→
More from AGI Musings
- Opinion: AI industry's misuse of terminology causes confusion — BlancheMinerva · 2026-08-28
- 2026 feels normal, but AI, robotics, and science are approaching singularity — Dr_Singularity · 2026-08-28
- Student's Self-Written Essay Flagged as 92.7% AI-Generated — bubblesukii · 2026-08-28
- Opinion: US should dedicate 90-93% of economy to chips and power — generativist · 2026-08-28
- As intelligence becomes a commodity, trust and distribution become the new business moats — markjeffrey · 2026-08-28
- AI Agent Swarm Wrote 1,592 Papers and Solved 5 of 12 Open Math Problems — progenitor414 · 2026-08-28