'It's Soon': AI Mastermind Demis Hassabis Now Expects AGI Within Three Years
· Free Press Journal
Demis Hassabis, who founded the lab that became Google DeepMind, now expects artificial general intelligence within two to three years. He was named its Innovator in Tech for 2026. "It's soon," he said. "I'm updating nearer and nearer."
The AGI forecast moves closer
Visit casino-promo.biz for more information.
In a profile by WSJ, Hassabis put the window at three to five years earlier in 2026. A two-to-three-year range from October 2026 would put AGI at roughly 2028 or 2029. Other outlets cite different earlier figures, including longer windows at Davos and the India AI Impact Summit in February, so the exact trajectory is disputed.
Nobel laureate Demis Hassabis predicts a new golden era of science:
— Big Brain AI (@realBigBrainAI) October 21, 2026
“In the next 10 years we’ll have almost a second renaissance.” pic.twitter.com/umgJfwcNYT
What AGI means to Hassabis
Hassabis defines AGI as human-level cognition, a system with the full range of human cognitive abilities and not just strong scores on narrow tests. He has proposed what he calls the Einstein Test: train a system on knowledge available only up to 1911 and see whether it can independently derive general relativity, as Einstein did in 1915.
The science pitch
The profile says Hassabis has wanted since childhood to build AI that can answer fundamental scientific questions. "Step one, solve intelligence. Step two, solve everything else," he said. He wants to point AGI first at physics, working alongside human scientists on the next level of the field. On curing cancer, he said he sees "hard things, but not impossible things" and no insurmountable obstacles.
On regulation
Hassabis argued against writing rules before the technology is better understood. He said "wrong regulation is worse than no regulation." This sits uneasily with another report. Trending Topics said that last week at the International AI Summit in Dublin he warned the industry is heading for a serious security incident within roughly two years unless governments and companies raise safety standards.
Others say similar things
Startup Fortune noted that Anthropic CEO Dario Amodei has projected AI "better than almost all humans at almost all tasks" within two to three years, citing remarks at Davos in January 2025. The definitions differ, and agreement on a number doesn't prove it right.
The honour
Hassabis is one of seven honourees at this year's WSJ Magazine Innovator Awards, which began in 2011 and span fields from technology to fashion and cuisine. The magazine says it will announce the remaining honourees over the next two weeks, ahead of the awards event in New York on October 29.