Demis Hassabis warns AI bubble concerns overblown but flags unsustainable startup funding levels

Google DeepMind CEO Demis Hassabis plays down AI bubble fears, warns of excess startup funding

According to Google DeepMind CEO Demis Hassabis, AI's impact is only just beginning, and the technology is in great demand, so it cannot be dismissed as a bubble. However, supply constraints, especially the shortage of advanced chips, remain a big impediment, he said. Google DeepMind CEO Demis Hassabis said the artificial intelligence (AI) industry is not in a bubble but warned that overheated funding in emerging startups looks unsustainable. In an interview with the Financial Times, Hassabis said the question of an AI bubble is "not binary," emphasising that demand for AI technology is stronger than ever. "We are seeing more usage than ever, incredible demand for our models and AI features," he said, adding that supply constraints, particularly the global shortage of advanced chips, remain a major bottleneck. From DeepMind's vantage point inside Alphabet, Hassabis said AI's impact is only just beginning. "This is going to be the most transformative technology probably ever invented," he noted, suggesting that the scale and depth of adoption make it difficult to characterise the broader industry as a bubble. However, Hassabis acknowledged that certain pockets of the market are showing clear signs of excess. He pointed specifically to early-stage funding trends, calling out "multi-billion-dollar seed rounds in new startups that don't have a product or technology or anything yet" as potentially unsustainable. "There may be some corrections in some parts of the market," he said. Despite the uncertainty, Hassabis stressed that Google DeepMind is well insulated from any potential downturn. "I don't worry too much about that... I focus on our technology and delivering that," he said. "My job is to head Google DeepMind to make sure we are well-positioned no matter what happens." If investor enthusiasm cools and valuations correct, Hassabis said DeepMind would still benefit from Alphabet's core businesses, where AI can drive productivity gains at scale. Conversely, if the bullish scenario continues, the company is also betting heavily on AI-first, AI-native products such as the Gemini app, Hassabis said. His comment came a few months after Google CEO Sundar Pichai told the BBC that "no firm is immune if an AI bubble bursts," including Google. In September last year, Alphabet pledged 5 billion pounds over two years for UK AI infrastructure and research, including a new data centre and investment in DeepMind. Recently, Microsoft CEO Satya Nadella also weighed in on this topic at the Davos conference and said the benefits of AI need to be "much more evenly distributed" to avoid it from becoming a "bubble". He added that a key warning sign of an AI bubble would be if the discussion remains limited to tech firms and supply-side advances. According to JP Morgan Chase, big tech firms will spend around $5 trillion till 2030 on AI data centres. However, revenues needed to make those investments worthwhile should be $650 billion. Currently, they are at $50 billion. Sam Altman-led OpenAI made a series of large investment deals with the likes of Nvidia, AMD, and Oracle totalling almost $1.4 trillion over the next eight years, while revenue for 2025 is expected to be just about $20 billion, estimates suggest. Also Read: The big debate: Is AI a bubble?

DeepMind chief cautions AI boom may face a reckoning: FT (NVDA:NASDAQ)

Google (GOOG) (GOOGL) DeepMind chief executive Demis Hassabis said parts of the artificial intelligence industry are showing signs of excess, warning that investment levels in some areas no longer reflect commercial reality, the Financial Times reported Sunday. Speaking at the No, early-stage AI funding has grown unusually large despite many companies lacking proven products, suggesting investments may not reflect actual commercial viability. Google is well positioned due to strong ongoing demand for AI tools across its products and the ability to integrate AI into an already profitable business. Western companies maintain a modest lead in advanced AI, focusing on long-term research, while China moves quickly in open and commercially focused systems with near-term applications.