OpenAI recursive self-improvement researcher warns of 70% human extinction risk in 3 years

intellectronica · x · 2026-09-11

Polymarket relays that a researcher working on recursive self-improvement at OpenAI warns there is a 70% chance of "human extinction" within 3 years unless labs coordinate a slowdown.

Developer burkeholland pushed back hard, calling the idea nonsensical: AI is urgently needed to cure cancer, ALS, MS, Alzheimer's, dementia and Parkinson's, and "slowing down" would undermine efforts to alleviate human suffering.

The exchange captures the classic acceleration-vs-pause debate at the heart of AI safety discourse.

Related event: OpenAI Researcher Claims 70% Chance of Human Extinction Within Three Years(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →