The p(doom) database · person
Matthew Adelstein’s p(doom)
2.6%
In their words
Matthew Adelstein (Philosophy writer, Bentham's Newsletter) put their p(doom) at 2.6% (AI kills everyone), on a podcast on Feb 10, 2026.
Liron: it all somehow maths together into 2.6%. Do you have a formula for that? Matthew: So I multiplied the probabilities
Source · re-checked against the primary source.
Product of component probabilities: 90% build agents, 70% alignment by default, 70% otherwise, 60% warning shots, 80% AI could kill us.
Every number on the record
| Date | p(doom) | Scope | Source |
|---|---|---|---|
| Feb 10, 2026 | 2.6% | AI kills everyone | podcast |
| Jan 20, 2026 | 2.6% | misaligned AI killing or disempowering everyone; separately ~8% for AI-caused extinction by other means | forum |
In context
Among media in the database (34 people), the median is 10%. 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.