Google DeepMind chief Demis Hassabis says there’s a coin‑flip chance that artificial general intelligence will arrive within five years.
What Happened: “My estimate is sort of 50% chance by in the next five years, so by 2030 let’s say. So I think there’s a good chance that that could happen,” he told the Lex Fridman Podcast, adding that his threshold for “true” AGI remains high.
Hassabis described AGI as intelligence that matches the cognitive functions of the brain, not the “jagged” profiles seen in today’s systems. “It isn’t kind of a jagged intelligence where some things, it’s really good at … but other things it’s really flawed at … You’d want that consistency of intelligence across the board,” he said.
He added that missing capabilities include “true invention capabilities and creativity.”
Hassabis proposed stress‑testing a candidate system on “tens of thousands of cognitive tasks that we know that humans can do” and inviting “a few hundred of the world’s top experts” to probe for failures over a month or two. If those experts find none, he said, “you can be pretty confident we have a fully general system.”
See Also: Mark Zuckerberg’s Meta Platforms Dodges $8 Billion Trial With Eleventh-Hour Settlement
The definition of AGI remains fluid. Tom’s Guide notes that some experts define AGI as human‑level competence across all domains, while others focus on its capacity to learn, adapt, and perform tasks that mirror human cognition, which requires the system to go beyond its initial training data and generate truly autonomous output.
Why It Matters: Hassabis has recently sketched different timelines, telling one interviewer he expects AGI “just after 2030” and emphasizing, “I have quite a high bar.” His stance contrasts with Google co‑founder Sergey Brin’s prediction of AGI before 2030, a split that reflects the field’s uncertainty over both definitions and pace.
Other leaders offer competing clocks. Anthropic’s co‑founder has said AGI is “possible” by 2028, while many researchers debate whether AGI should mean human‑level performance across most tasks. Meanwhile, OpenAI CEO Sam Altman thinks that AGI could arrive during Donald Trump’s presidency.
Meanwhile, former Google CEO Eric Schmidt, at a fireside chat in April, mentioned that it would be safe to assume that AGI would likely be achieved anytime between 2028 and 2030.
Photo Courtesy: khunkornStudio on Shutterstock.com
Read Next: