Slow-takeoff champion Paul Christiano now says fast takeoff may be months to years away
DavidSKrueger · x · 2026-09-10
Paul Christiano, arguably the most prominent advocate of slow AI takeoff over the past decade, now believes we're plausibly only "several months to several years" away from a fast takeoff. The shift comes as he announces he is joining OpenAI's nonprofit board to serve on the Safety and Security Committee, citing the recent trajectory of model capabilities.
More from AGI Musings
- Ex-OpenAI/Anthropic Researcher's Viral Resignation Post Suspected to Be a Psyop — TinfoilTricorn · 2026-09-10
- Where are the Chinese (p)doomers? AI doom beliefs may be a uniquely Western phenomenon — NinaDSchick · 2026-09-10
- 'AI Safety Has Killed More People Than AI': Accelerationists and Safetyists Clash — curious_vii · 2026-09-10
- Sentdex calls out AI fearmongering: those selling the fix are manufacturing the problem — Sentdex · 2026-09-10
- Doomer panic from big AI labs correlates directly with rising open-weight competitors — elmanmansimov · 2026-09-10
- Five Doctors Missed It in 3 Years — One AI Query Could Have: A Call for Volunteer AI Outreach — joshwhiton · 2026-09-10