The p(doom) database · person
Daniel Kokotajlo’s p(doom)
15% (if Plan A)
In their words
Daniel Kokotajlo (Executive director, AI Futures Project) put their p(doom) at 15% (if Plan A) (total catastrophe, conditional on the world adopting AI Futures Project's Plan A), on a podcast on Aug 27, 2026.
I think we say something like 15% chance of total catastrophe, conditional on doing Plan A. Yeah, even in Plan A it's like 15%.
Source · re-checked against the primary source.
Conditional estimate; he puts only 5–20% on Plan A actually happening.
Every number on the record
| Date | p(doom) | Scope | Source |
|---|---|---|---|
| Aug 27, 2026 | 15% (if Plan A) | total catastrophe, conditional on the world adopting AI Futures Project's Plan A | podcast |
| Oct 20, 2025 | 70–75% | bad outcomes incl. AI takeover, extinction and irreversible power concentration (stated as 25–30% chance of a good future) | podcast |
| May 25, 2023 | 70% | AI doom (unaligned superintelligence catastrophe) | forum |
In context
Among ex-frontier labs in the database (8 people), the median is 12.5%. Across all 223 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.