AGI is coming within four years and could start a new human era, says Google DeepMind CEO

Published 10 June 2026 · tech

AGI is coming. But when? If Google DeepMind CEO and Nobel laureate Demis Hassabis is right, the answer could be closer than many expect: around

AGI is coming. But when? If Google DeepMind CEO and Nobel laureate Demis Hassabis is right, the answer could be closer than many expect: around 2030. And when it does arrive, it could change the world we live in. According to Hassabis, the AI breakthroughs we are seeing today are only the beginning. He believes that once artificial general intelligence (AGI) arrives — AI capable of matching or surpassing human abilities across a wide range of tasks — humanity will enter an entirely new era. Read Full Story Speaking during a fireside chat at Stanford University's Graduate School of Business, Hassabis said AGI may arrive by "2030, plus or minus a year". He described the technology as so transformative that its arrival could feel like the start of a "new human era". According to Hassabis, society is currently standing in the "foothills of the singularity", the early stages of a technological shift that could fundamentally reshape how people live and work.

While the idea has long been discussed in science fiction and tech circles, he believes it is no longer a distant concept but something that is starting to take shape today. "We don't have long to prepare," he said, adding that the next few years will be critical in determining how humanity chooses to use the technology. The comments come from someone who has spent decades pursuing the idea of building intelligent machines. Looking back at DeepMind's beginnings in 2010, Hassabis said the company's mission was surprisingly simple: "Step one, solve intelligence. Step two, use it to solve everything else." According to Hassabis, that mission has largely remained unchanged. The goal was always to build AGI and then use it to tackle some of the biggest scientific challenges facing humanity, particularly in medicine and scientific discovery. Looking back at where it all started, Hassabis recalled how DeepMind's early AI systems struggled with even the simplest tasks.

One of the company's first breakthroughs came when an AI trained only on screen pixels eventually learned to play the classic game Pong. For weeks, it could not even score a point. Then it began winning games, convincing researchers that their approach had real potential. That progress eventually led to AlphaGo, which defeated world champion Lee Sedol in 2016. For Hassabis, the victory marked a turning point. It was proof that AI could generate genuinely new ideas and strategies that humans had never seen before. The second big breakthrough was AlphaFold, DeepMind's AI system for predicting protein structures. For Hassabis, it was proof that AI could do more than excel at games, it could help solve real-world scientific problems. He described protein folding as one of biology's biggest unsolved challenges. The impact of AlphaFold was so significant that Hassabis, alongside John Jumper and David Baker, was awarded the 2024 Nobel Prize in Chemistry for their contributions to protein structure prediction.