Google DeepMind CEO Demis Hassabis has warned that parts of the booming artificial intelligence investment landscape may be about to correct, implying that early-stage AI startups are overvalued and aimed at the growing investor frenzy.
During a recent episode of Google DeepMind: The Podcast, Hassabis stated that he sees signs of speculative excess, particularly among young AI companies raising capital at extremely high valuations despite making little progress or gaining traction. He also questioned the long-term viability of startups that receive multibillion-dollar valuations at the seed or early growth stage, pointing out that many of them have only recently begun operations.
He did, however, draw an obvious line between these early-stage funding rounds and the massive investment made by established tech firms. He stated that Big Tech companies investing billions of dollars in AI infrastructure and research are backed by existing businesses, revenues and long-term strategic roadmaps, making their valuations more credible.
According to him, artificial intelligence is currently overhyped in the short term, while its long-term impact is underestimated. He also stated that major technological shifts frequently transition quickly from widespread scepticism to intense enthusiasm, resulting in inflated expectations and excessive valuations before the market adjusts. Such cycles, he added, usually result in an ‘over-correction’ as reality catches up to hype.
When discussing DeepMind’s early days, Hassabis stated that AI research struggled to attract belief or funding when the company was founded more than a decade ago. In contrast, artificial intelligence (AI) now dominates global business, investment and policy discussions. He suggested that the sharp swing in sentiment has contributed to some parts of the market moving too far, too quickly.
Despite these concerns, Hassabis stated that he is not concerned with whether AI is in a bubble. His primary focus remains on developing foundational AI systems at Google DeepMind, which creates models that power Google products like Gemini and oversees the company’s frontier AI research.