Demis Hassabis
Co-founder and CEO of Google DeepMind
Chess prodigy, game developer and neuroscientist who co-founded DeepMind in 2010. He shared the 2024 Nobel Prize in Chemistry for AlphaFold and leads Google's combined AI research organization.
Read them in their own words
Published elsewhere, not by me — so you get real context beyond the name and the title.
Demis Hassabis Is Preparing for AI's Endgame
Traces the line from a Nobel-winning protein-folding model to his stated goal of building general AI.
Key arguments & positions
- Frames AI primarily as an instrument for scientific discovery — protein structures, materials, weather, mathematics.
- Supports international coordination on frontier AI testing, and has signed statements putting extinction-level risk on the global agenda.
- Has said scaling alone is unlikely to be enough for general intelligence, and that new research ideas are still needed.
Accomplishments
- Co-founded DeepMind in 2010; CEO of Google DeepMind since the 2023 merger with Google Brain.
- Shared the 2024 Nobel Prize in Chemistry for AlphaFold's protein-structure predictions.
- Oversaw AlphaGo, AlphaZero, AlphaFold and GraphCast; knighted in 2024.
Papers & key writings
Links
Jobs
What they've said in public about careers, hiring and work. Each point links to its source.
- Public position
Said AI will be a bigger transformation than the Industrial Revolution, but was optimistic because humans are "infinitely adaptable." He expects "new jobs will appear, new very valuable jobs," and that AI will "supercharge sort of technically savvy people who are at the forefront of using these technologies."
Business Insider (SXSW London talk), June 2025 - Career advice
Told Cambridge students to use their time as undergraduates to understand how they learn best — "learning to learn" and adaptability matter more than any single subject, because you will need to pick up new material quickly throughout your career.
Business Insider (Queens' College, Cambridge interview), May 2025
In the library
Nothing of theirs is filed in the hub yet. Browse the full library.
Recommended next
Hand-picked from the hub based on what Demis Hassabis covers.
A proof of the Aaronson–Ambainis conjecture
PSI reports that its internal research system, using a non-public Anthropic model, produced a proof of the Aaronson–Ambainis conjecture and a Lean formalization. The announcement explains the claimed implications for quantum query speedups and links to a paper and proof repository.
Why this: Also about AI for science
Training Small Language Models for Transferable Tool Selection
Preprint, not yet peer-reviewed. Researchers at Meta and Carnegie Mellon University study reinforcement learning for tool selection by a small language model and transfer to different retrievers and tool collections. The authors report that their post-trained Qwen3-8B outperforms Llama 4 Maverick on ToolBench; this is the paper’s reported result, not an independent benchmark verification.
Why this: Also about reinforcement learning
PSI · Physical Superintelligence
The site of PSI, a company describing itself as an AI-native laboratory applying AI to physics problems in energy, computation, materials and sensing.
Why this: Also about AI for science
Investing in Preference Model
a16z post from October 7, 2026 announcing an investment, amount not disclosed, in Preference Model, a company that builds reinforcement-learning environments for AI labs. The post says the company is open-sourcing a framework called Karotte.
Why this: Also about reinforcement learning