According to Financial Times News, Google DeepMind chief Demis Hassabis warned at the World Economic Forum in Davos that AI investment is looking “bubble-like,” citing “multibillion-dollar seed rounds” for startups without a product or technology. He pointed to examples like the Thinking Machine Lab, valued at $10bn just six months after founding, as unsustainable and predicted potential “corrections.” Hassabis argued Google, with its parent Alphabet now valued over $4tn, would be fine if a bubble bursts because it can add AI features to its existing business. He also claimed Western companies, led by Google’s Gemini 3 model, still hold a “six months or so” lead over Chinese AI developers like DeepSeek, despite their advances in open models. This comes as other Davos leaders like Nvidia’s Jensen Huang dismissed over-investment concerns.
Bubble Talk and Big Tech Bunkers
Here’s the thing: when the guy running one of the most expensive AI research labs on the planet starts talking about a bubble, you should probably listen. But his argument is fascinatingly self-serving. Sure, those $10bn valuations for a six-month-old startup with no product seem insane. Everyone with a brain knows that. But Hassabis is basically saying, “That wild speculation over there is a problem. Our wild speculation over here, embedded in a trillion-dollar ad business, is totally fine.” He’s not wrong about Google’s defensive position. If the froth evaporates, they have a massive, profitable core business to retreat into. The startups betting everything on a single, unproven model? Not so much. It’s a classic case of “it’s a bubble, but I’m in the castle, and you’re in the straw house.”
innovation-claims”>The China Factor and Innovation Claims
His comments on China are the real eyebrow-raiser. Dismissing DeepSeek’s powerful, open, and cheap model as something the West “overreacted” to feels like dangerous complacency. He says Chinese labs haven’t proven they can innovate “beyond the frontier yet.” Okay, but what if the near-term game isn’t about leaping the frontier, but dominating the practical, applied AI that gets built into everything? Hassabis admits they’re focused on near-term applications for revenue. That sounds… like a good business strategy? The West’s obsession with the moonshot toward Artificial General Intelligence (AGI) is noble, but it’s also a bottomless pit for capital. China flooding the market with capable, free-to-use models could reshape the entire developer ecosystem, making the “six-month lead” irrelevant if everyone starts building on their stack. It’s a different kind of race.
Smart Glasses Again, Really?
And then we get to the real fun part: the return of smart glasses. Google pushing this again feels like a tech industry trauma response. They failed spectacularly with Google Glass over a decade ago. Now, Hassabis thinks an AI “universal digital assistant” could be the killer app they lacked. I want to believe, but the hardware challenges and social acceptance issues haven’t magically disappeared. It’s a telling pivot, though. It shows that even at DeepMind, the pressure is on to find tangible, consumer-facing applications for this world-changing tech. Science and medicine are “unequivocal goods,” but they don’t drive quarterly earnings in the same way. The scramble to justify the spending is leading them back to some very old, very dusty ideas.
The Unsustainable Pace
Maybe the most human quote in the whole piece is Hassabis dismissing the idea of ever becoming Alphabet CEO. “There’s only so much one can do in the day and still leave enough time for serious thinking,” he said. That, right there, might be the most profound warning about the current AI boom. The breakneck speed, the insane funding rounds, the hype cycle—it’s actively hostile to “serious thinking.” And if we’re building a technology that the chief architect himself calls “the most transformative probably ever invented,” shouldn’t we want the people in charge to have time for serious thought? The bubble isn’t just about money. It’s about the unsustainable cognitive and cultural pace required to keep up with it. That’s a correction no company, not even Google, is immune to.
