Jan 23, 2026 · 33m · big-technology
Google DeepMind CEO Demis Hassabis: AI's Next Breakthroughs, AGI Timeline, Google's AI Glasses Bet
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this Davos edition of the Big Technology Podcast, Google DeepMind CEO Demis Hassabis discusses the realistic path and architectural breakthroughs required to reach AGI, Google's roadmap for AI-powered smart glasses, and the transformative potential of AI in scientific discovery.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Alex holds 23.3% of the talking time here. How this is scored →
Alex as informed peer 4.3 Guest teaching 4.4 Guest disagreement 1.6 Alex pushing back 3.1
speaking balance: gold is Alex, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 6:56 Rejection of Altman's AGI marketing definition
Hassabis bluntly dismisses Sam Altman's attempt to redefine AGI as commercial marketing, insisting on strict scientific benchmarks.
Hardest push from Alex ▶ 9:54 Host challenges Nano Banana AGI claimKantrowitz expresses open disbelief at Hassabis claiming an image generator is close to AGI, forcing an immediate explanation.
Biggest teaching moment ▶ 10:04 Hassabis educates on generative world modelsHassabis explains how video generation represents an intuitive physics world model that provides the foundation for real-world planning in AGI.
Alex holds their own ▶ 20:11 Host's three-step AI industry collapse frameworkKantrowitz demonstrates deep market knowledge by presenting a structured three-step bear case linking diminishing training returns, cheap flash models, and capex overhang.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Alex as informed peer | Guest teaching | Guest disagreement | Alex pushing back | Why |
|---|---|---|---|---|---|---|
| Evaluating AI Progress and Overcoming the Scaling Wall | 5 | 4 | 2 | 4 | Kantrowitz articulates a sharp skeptic argument regarding AI memory and scaffolding. Hassabis counters the scaling wall narrative, explaining that pre-training, post-training, and synthetic data still offer significant headroom. | |
| Neurosymbolic AI Systems and the Continual Learning Challenge | 4 | 5 | 1 | 3 | Kantrowitz asks whether hard-coded systems can be AGI and presses on the continual learning problem. Hassabis breaks down neurosymbolic hybrid architectures like AlphaFold and AlphaGo. | |
| Scientific Definitions of AGI Versus Superintelligence Timelines | 4 | 6 | 6 | 3 | Kantrowitz cites Sam Altman's claim that AGI is underdefined and already passed. Hassabis forcefully dismisses Altman's view as commercial marketing and outlines rigorous scientific criteria like Einstein-level scientific discovery. | |
| Generative Video Models and Physical World Simulation | 4 | 6 | 2 | 5 | Kantrowitz directly expresses incredulity at Hassabis citing an image generator as near-AGI. Hassabis explains how generative video models build intuitive physical world models essential for long-term planning. | |
| The Roadmap for AI Smart Glasses and Digital Assistants | 4 | 2 | 0 | 2 | Kantrowitz highlights the awkwardness of phone-based assistants in DeepMind's documentary and asks about smart glasses. Hassabis agrees on the form factor, revealing DeepMind's partnerships and summer rollout timeline. | |
| Evaluating AI Chatbot Business Models and User Trust | 5 | 3 | 2 | 5 | Kantrowitz repeatedly presses on whether Google plans to insert advertising into Gemini. Hassabis explains the fundamental tension between advertising and user trust before giving a definitive denial of current ad plans. | |
| The Coding Revolution: Claude Code and Google Anti-Gravity | 6 | 4 | 1 | 4 | Kantrowitz lays out an elaborate three-step thesis on how AI infrastructure spending could trigger a market collapse. Hassabis acknowledges the bubble risks among early-stage startups while defending Alphabet's dual positioning. | |
| Human Adaptation, Labor Automation, and Finding Purpose Post-AGI | 4 | 5 | 1 | 2 | Kantrowitz questions whether human knowledge workers will end up demoralized like elite gamers beaten by AI. Hassabis counters with the history of chess and Go, arguing humans are adaptable toolmakers who will redefine purpose. | |
| Exploring Information as the Fundamental Unit of the Universe | 3 | 6 | 0 | 1 | Kantrowitz asks Hassabis to unpack his philosophical view that information is the fundamental unit of reality. Hassabis details biological entropy resistance and how AlphaFold navigates information topologies. | |
| AlphaFold's Global Open Release and DeepMind Research Culture | 4 | 3 | 1 | 2 | Kantrowitz asks about the decision to open-source AlphaFold and whether it mirrored internal tensions with Google corporate. Hassabis clarifies the scientific necessity of open access and draws parallels to future AlphaZero-style scientific discovery. |
Statements from this episode (23)
Hassabis: AI Has Significant Headroom Using Existing Architectures and Data
“Actually it turns out you can wring more more juice out of the existing Architectures and data. So there's plenty of room, I think, and we're still seeing that in both the pre-training, the post-training, and the thinking paradigms, and also the way that they …”
Hassabis: Reaching AGI Will Require One or Two More Fundamental Breakthroughs
“I'm definitely a subscriber to the idea that maybe we need one or two more big breakthroughs before we'll get to AGI, and I think they're along the lines of things like continual learning, better memory, longer context windows, or perhaps more efficient contex…”
Demis Hassabis Rejects Yann LeCun's Stance That LLMs Are A Dead End
“No matter what camp you're in, we're gonna need Large foundation models as the key component of the final AGI systems. Of that, I'm sure. So, I don't, I'm not subscriber to someone like Jan LeCun who thinks, you know, that there's sort of some kind of dead end…”
Hassabis: General Learning Across Any Domain Is The Defining Feature of AGI
“Learning is a critical part of agenda of AGI. It's actually almost the defining feature. When we say general, we mean general learning. Can it learn new knowledge and can it learn across any domain? That's the general part. So for me, learning is synonymous wi…”
Hassabis: Continual learning that updates models over time has not been cracked
“But I think to have it you want to do it more than just having your data in the context window. That's you want to have something a bit deeper than that, which is, as you say, actually changes the model over time. That's what ideally you would have. And that t…”
Hassabis: AGI Must Mean Matching All Human Cognitive Capabilities
“I don't think AGI should be sort of turned into a marketing term or for commercial gain. I think there is always been a scientific definition of that. My definition of that is a system that that can exhibit all the cognitive capabilities humans can. And I mean…”
Hassabis: Today's AI systems are nowhere near true human-level invention
“And today's systems, in my opinion, are nowhere near that. Doesn't matter how many, you know, Erdos problems you solve, which for some reason, you know, I mean, you know, that's, it's good that we're doing those things, but I think it's far, far from what you …”
Hassabis Predicts True AGI Is Five To Ten Years Away
“I think an AGI system would have to be able to do all of those things to really fulfill the original sort of goal of the AI field. And I think, you know, we're five to 10 years away from that.”
Hassabis: World models are essential for AGI and long-term planning
“And then, of course, that would be, I think, essential for a GI, because that would allow these systems to plan, long-term plan, in the real world over, perhaps, very long time horizons, which, of course, we as humans can do, you know I'll spend four years get…”
DeepMind Plans to Converge All Modalities Into One Unified Gemini Model
“Why this is why we worked with Gemini as being multimodal from the beginning, able to deal with, you know, video, image and eventually converge that all into one model, that's our plan, is that it'll be very useful for a universal assistant as well.”
Next-Generation Google and Samsung Smart Glasses Will Appear By Summer 2026
“We've done some great partnerships with Warby Parker and Gentle Monster and Samsung to build these next generation glasses. And you should start seeing that you know, maybe by the summer.”
Google Has No Current Plans to Introduce Ads to the Gemini App
“From our point of view, we have no plans at the moment to do ads. If you're talking about the Gemini app, right? Specifically”
Hassabis: Vibe Coding Will Enable Creatives to Build Software Independently
“I think it will open up the whole productivity space to designers, creatives, artists that maybe would have had to work with teams about access to teams of programmers. Now they can probably do, you know, a lot more just on their own.”
Google Cannot Meet User Demand for New Anti-Gravity IDE
“We've just released anti-gravity, our own IDE which is very, very popular. We can't actually serve all the demand that we're seeing there.”
Hassabis: Anthropic Avoids Multimodal Models to Focus Strictly on Coding
“You know, they don't make image models, multimodal models, world models. They just do, you know, coding and language models, and they're very, very good at that.”
Hassabis: Today's AI models have a huge unknown capability overhang
“I think there's a huge capability overhang actually of what even today's models can do that maybe even us as building those things don't fully know.”
Demis Hassabis: Multi-Billion Dollar AI Seed Rounds Are Unsustainable and "Filthy"
“When you see seed rounds of tens of billions of dollars of companies that basically have no product or research, it's just Some people coming together. That seems a bit unsustainable to me in a normal market, a bit filthy.”
Hassabis: Top Go player reached record Elo learning from AlphaGo
“The best South Go player in the world is a South Korean, and he was about 15, I think, when AlphaGo match happened. He's in his mid-twenties now, and he's by far the strongest player there's ever, Been by the ELO ratings, because he's learned natively, young e…”
Hassabis: AI automation may be 10x the Industrial Revolution
“I think we've, it's like the industrial revolution, maybe 10 X of that, but we'll have to adapt again, and I think we'll find new meaning in things, and we do a lot of things already today that are not just for economic gain.”
Hassabis: Three million researchers worldwide use AlphaFold
“And it's been incredibly gratifying to see, you know, three million researchers around the world use it in their important research.”
Hassabis: Almost every future drug discovered will probably use AlphaFold
“I think in future Almost every single drug that's discovered from now on will probably have used AlphaFold at some point in that process, which is, you know, amazing for us, and, you know, really, that's what we do, all the work we do for.”
Hassabis: Post-human AGI will discover room-temperature superconductors and new energy sources
“I mean, that's what to me is, it would be the AGI moment is, you know, then it will discover a new superconductor, room temperature superconductor that's possible in the laws of physics, but we just haven't found that needle in the haystack. Or a new source of…”