AI will not undergo a "hard take-off" (an AGI recursively self-improving into vastly superhuman intelligence within hours or days); no window stated, inferred window about 10 years (to April 2033).

Pending ●●●●●

“The "hard take-off" scenario is utterly impossible.”

Said on: 25 April 2023 · Post on X replying to Perry Metzger's question about Eliezer Yudkowsky's "foom" scenario. No window stated; long-run claim, so inferred window is about 10 years.
Source: It seems AI doomers need it to be spelled out · X (@ylecun)
Should come true: 2030s
How sure they sounded: utterly impossible
Categories: AI
Importance: 4 of 5 · A sudden intelligence explosion would have major global consequences; the claim is central to AI-risk debates.

Outcome

Not scored yet.

History