The p(doom) database · person
Roko Mijic’s p(doom)
~35%
In their words
Roko Mijic (Rationalist writer (Roko's basilisk)) put their p(doom) at ~35% (AI disaster ("a full-stack takeover of human society")), on X on Oct 28, 2026.
AI disaster is going to look like a full-stack takeover of human society as we hand over more and more of it to machines. My P(Doom) is like 35%
Source · re-checked against the primary source.
Unconditional. His Oct 18 "95%+" was "on the current path"; this is back in his 35–40% range.
Every number on the record
| Date | p(doom) | Scope | Source |
|---|---|---|---|
| Oct 28, 2026 | ~35% | AI disaster ("a full-stack takeover of human society") | x |
| Oct 18, 2026 | 95%+ (on the current path) | the end of the world from AI · on the current path | x |
| Sep 26, 2026 | 35–40% | p(doom) | x |
| Jan 30, 2024 | 40–50% | p(doom) | x |
| Jan 19, 2024 | ~50% | p(doom), mostly driven by misuse of ASI by insiders | x |
In context
Among independent in the database (68 people), the median is 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.