Geoffrey Hinton’s p(doom)
In their words
Geoffrey Hinton (Nobel laureate; University of Toronto) put their p(doom) at 10–20% (AI tries to take over from us), (tv) on Oct 14, 2026.
That's the point of saying 10 percent to 20 percent, to tell people it's a significant chance. It's not inevitable. That means it's -- that's saying there's an 80 percent chance it won't happen.
Source · re-checked against the primary source.
CNN The Situation Room, answering Jensen Huang. Says "Nobody can get a good number. There's no way to justify 10 percent or 20 percent. You use those numbers to indicate to people it's a significant chance." Back to his usual range after 10% in September. Oct 21, El País: "It's more than 1%, so we ought to really worry about it" (a lower bound; headlines misquoted it as 1%). Starmus, Tenerife (~Oct 20, La Vanguardia, Oct 25): above 1% and below 99%; "10% or 20%, which is what I've been saying, is very worrying."
Every number on the record
| Date | p(doom) | Scope | Source |
|---|---|---|---|
| Oct 14, 2026 | 10–20% | AI tries to take over from us | tv |
| Sep 9, 2026 | 10% | AI kills all humans · within a decade | interview |
| Dec 27, 2024 | 10–20% | AI catastrophe / extinction · within 30 years | article |
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.