"Against the AI Omnicause": New Essay Says Safety Orgs Dodge Extinction Risk

zetalyrae · x · 2026-09-30

A new Cognition Café essay by Gabe and Fernando Borretti (co-worked on by zetalyrae) argues AI safety organizations privately believe extinction is the main risk — CEOs themselves put "really, really badly" at 1-in-4 — yet publicly advocate an ever-growing "omnicause" list of everything but extinction, from deepfakes to job loss. The authors compare it to debating third-class gym access on the Titanic, and conclude that if nothing else suffices, we should work directly to ban ASI.

Original post →

More from AGI Musings

AGI Musings channel →