Why people stay at AI labs despite believing in ~10% extinction risk: three rationales and a void of meaning

sebkrier · x · 2026-09-10

Jasmine Sun lays out three reasons people keep working at AI labs while believing extinction risk is around 10%: techno-determinism (someone will build ASI anyway, better I do it safely), consequentialism (utopia vs doom makes it a +EV bet), and self-interest (fun, money, friends). She notes nobody admits hyping risk for marketing—people genuinely believe it while compartmentalizing. Quoted scholar ScholarsStage adds a rarely articulated driver: the work fills a void of meaning, echoing the psychology of past revolutionaries who saw themselves as agents of history.

Related event: Why AI Researchers Stay Despite Believing Extinction Risk(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →