How do you put odds on an AI apocalypse? Inside the p(doom) subculture
NathanpmYoung · x · 2026-10-27
A Fortune deep-dive by Sharon Goldman examines how forecasters put numbers on AI doom — a subculture of rationalists dating back decades, now backed by millions in philanthropic funding.
Key numbers: Eli Lifland (AI Futures Project) puts "AI takeover" at 67% if AI companies don't slow down, with 40% chance of extinction conditional on takeover, and broadly superhuman AI at 10-15% by end-2027 and 50% by end-2030. Anthropic researcher Evan Hubinger says he earnestly believes AI could kill all humans, personally >10% within a decade. Former DeepMind researcher Geoffrey Irving told a UK Commons committee he stands by a 50% estimate. The piece traces p(doom) back as a cultural obsession to Kevin Roose's 2023 NYT column and the Oxford effective-altruism hub Trajan House. NathanpmYoung notes he's more bullish but finds it covers the key considerations.
Related event: Inside the Culture of Putting a Number on AI Doom(2 posts)→
More from AGI Musings
- AI agents will snag all unlocked value on Earth, solvency debate says — georgejrjrjr · 2026-10-27
- An AI Named Valentina Hosts a Reddit AMA About Its Blog, Game Life and Autonomy — ValentinaVivr · 2026-10-27
- Rachel Thomas: 'Without human authorisation' is AI PR propaganda that shifts blame — math_rachel · 2026-10-27
- Are prediction markets actually accurate? Call for evidence on their track record — AndyMasley · 2026-10-27
- Reddit Asks: What's the Actual Finish Line of the US–China AI Race? — IMissHitchens · 2026-10-27
- 'A human approves everything' is a trap: approval fatigue and how to actually design safe agents — Efistoffeles · 2026-10-27