According to Bloomberg Business, Google plans to launch AI-powered smart glasses in 2026, developed in partnership with Samsung. This marks Google’s second attempt, following the infamous Google Glass, and is a direct challenge to Meta. The project is deeply tied to Demis Hassabis, the head of Google’s merged AI units, who believes “world models” trained on the physical world are the next paradigm, not just chatbots. His leadership was solidified in 2023 when CEO Sundar Pichai put him in charge, and he later managed the 2024 return of Transformer co-inventor Noam Shazeer. The pressure is on: while Hassabis’s Gemini 3 now has over 650 million monthly app users, Google needs to monetize AI beyond ads, and Hassabis needs a product breakthrough after years of pure science.
Hassabis Versus the Chatbot World
Here’s the thing: the entire AI industry has been sprinting down one path for the last few years. It’s the path of the large language model, the chatbot, the thing that powers Gemini and ChatGPT. Billions of dollars and insane compute power, all aimed at predicting the next word. Hassabis has never fully bought it. He’s been the guy in the room arguing for “world models”—AI that understands physics, space, and how objects relate. Think of it as common sense for machines.
And that’s why these glasses matter so much. They’re not just a wearable screen for notifications. They’re the potential vessel for his Project Astra vision: AI that sees and comprehends the 3D world you’re in. It could remember where you left your keys, not just describe a picture of a keychain. That’s a fundamentally different ambition than Meta’s Ray-Bans, which are basically fancy camera-connected speakers. If Hassabis is right, chatbots are just a phase, not the destination.
The Chess Prodigy’s Gamble
But let’s be real. Hassabis’s track record is… academic. Prestigious? Absolutely. AlphaFold is a Nobel-winning miracle for biology. But as Bloomberg notes, it hasn’t created a single FDA-approved drug yet. Beating Go champions is incredible for PR, but it doesn’t pay Google’s cloud bills. He’s been the brilliant scientist giving Google a halo effect. Now, he needs to be the product architect.
That’s why the corporate maneuvering is so fascinating. Pichai betting on him over a more traditional engineering lead like Jeff Dean. Hassabis, from London, somehow smoothing over Google’s internal AI wars. Then, the masterstroke of re-hiring the mercurial Noam Shazeer—the chatbot true believer—and making it work through “charm and diplomacy.” That’s not luck. That’s a chess player thinking several moves ahead, consolidating his pieces for the endgame. The endgame being 2026.
The Real Killer App?
So, can smart glasses actually be the killer app? I’m skeptical, and you probably are too. Google Glass left a permanent stain on the category. The tech has to be invisible, stylish, and undeniably useful from day one. But the potential is there. An AI that truly understands your physical environment is a leap into a new computing paradigm. It’s not about asking a bot to write an email; it’s about an assistant that exists in your world.
For Google, the stakes are existential. They’re playing catch-up in the AI narrative, despite their massive user base. Ads won’t fund the AI future alone. They need a hardware-software breakthrough that people will buy. For Hassabis, it’s legacy. Does he remain a decorated scientist, like a modern-day “AI architect” in the history books? Or does he become the person who steered Google into its next era, proving that his scientific vision was the commercially viable one all along?
The race isn’t just against OpenAI anymore. It’s Hassabis’s theory of AI against the prevailing dogma. And in 2026, we’ll all get to see if the glasses are a looking glass into the future, or just another high-tech hallucination.
