The p(doom) database · person

Paul Christiano’s p(doom)

~50% ("a coin flip")
Board member, OpenAI Foundation; founder, Alignment Research Center · ex-frontier labs · Nov 9, 2026

In their words

Paul Christiano (Board member, OpenAI Foundation; founder, Alignment Research Center) put their p(doom) at ~50% ("a coin flip") (AI takeover, if AI continues on "the most rapid trajectory"), (speech) on Nov 9, 2026.

So my feeling is if we try to develop AI along the most rapid trajectory, if we said we're just going to build this as quickly as we can in the easiest way we can, we would be looking at something like a coin flip of an AI takeover. I hope the probability is much lower than that because we are not going to develop the technology in the very fastest way we can. But I think even all things considered, with realistic policy responses, the risk remains exceptionally high.

Source · re-checked against the primary source.

Hinton Lectures 2026, night 1, Toronto (video: youtube.com/watch?v=-Ef0ocEoBCs; also quoted by Ars Technica). Conditional on the fastest trajectory; gives no unconditional number ("exceptionally high") and says no "quantitative model" can "put a number on this". Adds: "for most people in this room, the most likely reason they will die is because of an AI accident."

Every number on the record

Datep(doom)ScopeSource
Nov 9, 2026~50% ("a coin flip")AI takeover · if AI continues on "the most rapid trajectory"speech
Sep 9, 20264% / 15%catastrophic and irreversible loss of control · 1 year / 3 yearsarticle
Apr 27, 202346%irreversibly messed-up future (takeover 22% + extinction 9% + other 15%) · within 10 years of building powerful AIforum
Apr 21, 202310–20%AI takeover with many or most humans dead; separately 'maybe a 50/50 chance of catastrophe shortly after we have systems at the human level'podcast

In context

Among ex-frontier labs in the database (9 people), the median is 27.5%. Across all 252 people it is 10%. How the medians and the index are built.

A p(doom) is a personal, subjective estimate of the probability that advanced AI ends in human extinction or a comparable, irreversible catastrophe. Scope and timeframe differ from person to person, so compare with care: why p(doom) numbers disagree.

See all 346 statements in the database · everyone A–Z