Yoshua Bengio

“The duration of tasks AI models can solve doubles every seven months; in five years, AI will be at human level for programming tasks.”

Yoshua BengioResearcher

Professor, Université de Montréal; Founder and Scientific Director, Mila; Turing Award laureate

Deadline 2030; Bengio repeated the same reasoning in his December 2025 Digitalist Papers essay, putting AI planning ability at approximately human level around 2030 if the METR trend persists.

“Hyperintelligent machines could pose an existential threat to humanity within the next decade: AI systems trained on human language and behaviour could develop their own 'preservation goals', making them in effect competitors to the species that created them.”

Yoshua BengioResearcher

Turing Award laureate; professor, Université de Montréal; founder of Mila and LawZero

Bengio framed major AI risks as arriving in roughly five to 10 years while cautioning they could come sooner; he launched the nonprofit LawZero in June 2025 with $30 million to build 'non-agentic' safe-by-default systems.

“My current estimate places a 95% confidence interval for the time horizon of superhuman intelligence at 5 to 20 years.”

Yoshua BengioResearcher

Turing Award laureate; Professor, Université de Montréal; Scientific Director, Mila

The 95% interval runs to roughly 2043; the early end of the window (2028) will be the first informal checkpoint.