←News detail

The PaperUpdated Oct 25 21:07

Hassabis shortens AGI forecast from "three to five years" to "two to three years"

AI summary

In an interview with The Wall Street Journal, Google DeepMind CEO Demis Hassabis compressed the timeline for AGI's arrival from "three to five years" earlier this year to "two to three years," and said the expectation is still moving earlier. He attributed part of his judgment to Isomorphic Labs: the company completed a $2.1 billion Series B funding round in May 2026 and plans to submit its first new drug clinical trial application before the end of 2026. Hassabis reiterated that AGI's primary value is accelerating scientific discovery, and argued that rushing to legislate before the technology is fully understood means "bad regulation is worse than no regulation."

Body · AI translation

Download the app

+1 From "three to five years" to "two to three years": Hassabis predicts artificial general intelligence will arrive sooner

The Paper reporter Ji Jingjie 2026-10-22 11:13 Source: The Paper ∙ Future 2% > Listen to full text Font size

Google DeepMind CEO Demis Hassabis has given his most urgent forecast yet for the timeline of artificial general intelligence (AGI), which he has long tracked. In a recent exclusive interview, he further compressed the arrival of AGI from the "three to five years" he repeatedly mentioned earlier this year to "two to three years," and said outright that this judgment is continuing to accelerate.

Hassabis defines AGI as a system possessing all the cognitive abilities of the human mind. "Soon (it will appear)," Hassabis told The Wall Street Journal. "I'm updating the expected timeline to be closer and closer."

In an interview in April this year, he said "the probability of achieving AGI within the next five years is very high," and mentioned that computing power was the biggest bottleneck at the time. Two months later, in a conversation at Stanford Graduate School of Business, he narrowed the window to "three to five years, with an error margin of no more than one year," and rarely described the scale of this transformation in quantitative terms: "Its impact is 10 times that of the Industrial Revolution, at 10 times the speed—that is, completing a century of change in ten years."

This time Hassabis has moved this countdown further forward. One basis for the judgment comes from Isomorphic Labs, the drug discovery company he heads. The company was spun out of DeepMind in 2021 to commercialize AI biology breakthroughs such as AlphaFold. In May 2026, Isomorphic completed a $2.1 billion Series B financing led by Thrive Capital, bringing total external capital to about $2.7 billion, making it the largest financing to date in the field of AI-driven drug discovery.

According to The Pharma Letter, several of the company's drug candidates for oncology and immune-mediated diseases are in IND-enabling studies, and it plans to submit its first new drug clinical trial application and initiate first-in-human trials before the end of 2026. Its core platform IsoDDE claimed in a technical report released in February 2026 that on the hardest protein-ligand systems, its accuracy is more than twice that of AlphaFold 3.

Hassabis previously argued that the primary value of AGI is not automating everyday tasks but accelerating scientific discovery. He reiterated this logic in this interview: "The ultimate goal is to build AI and understand the mind, then use it to understand science. Step one, solve intelligence. Step two, solve everything else."

On regulatory issues, Hassabis's attitude continues his consistent cautious stance. He told The Wall Street Journal that rushing to legislate when the technology is not yet fully understood will almost inevitably produce wrong regulation, and "wrong regulation is worse than no regulation." He also stressed that discussing AI risks without proposing actionable solutions is not enough: "I think a lot about solutions. If there is a problem, how do we use humanity's collective intelligence to solve it?"

Hassabis was named one of the 2026 Innovators of the Year by The Wall Street Journal Magazine, one of seven awardees, with the award event scheduled for October 29 in New York. In the interview, he mentioned that if AGI appears, one of the first things he most wants to do is point it at fundamental problems in physics: "I hope that by then I can sit down with my AGI system and some great human scientists and try to figure out the next level of physics."

For the broader public, he believes the tangible feel of AI will first appear in the health field. "I think people may really start to feel the benefits of AI directly in their lives," he said. "I don't see any insurmountable obstacles. I see difficult things, but not impossible things."

Responsible editor: Huan Yanhong Image editor: Le Yufeng Proofreader: Zhang Yan The Paper news tip: 021-962866 The Paper, unauthorized reproduction prohibited +1 Favorite I want to report #Hassabis #DeepMind View more

Source

The Paper