Demis

Demis Hassabis founded DeepMind in 2010 and shared the 2024 Nobel Prize in Chemistry with John Jumper for their work developing AlphaFold, an AI system that predicts three-dimensional protein structures from amino acid sequences. David Baker was also awarded the prize for computational protein design. Hassabis is also the founder and CEO of Isomorphic Labs, where he pursues AI applications in drug discovery.[6][3][5]

AlphaFold addressed a complex, time-consuming problem: predicting a protein’s structure from its sequence. Its predictions were made freely available through the AlphaFold Protein Structure Database. Google DeepMind’s October 2024 award announcement reported that more than two million scientists and researchers in 190 countries had used them. In his accompanying statement, Hassabis thanked Jumper, the AlphaFold team, and colleagues across DeepMind and Google.[3]

Hassabis described the achievement as a starting point for broader scientific applications: “I hope we'll look back on AlphaFold as the first proof point of AI's incredible potential to accelerate scientific discovery.” His priority at Isomorphic Labs is explicit: “I believe there is no more important application for AI than helping to improve human health.” The company applies machine learning to drug discovery, building on and beyond AlphaFold.[3][5]

On August 5, 2026, Hassabis announced a leadership transition in a post reproduced by Latent.Space: “I’m stepping into a new role as Chair of Google DeepMind & Chief Scientist of Alphabet.” He said the change would allow him to focus on long-term strategy and accelerating scientific breakthroughs.[1]

Hassabis also publicly emphasized practical model performance. Latent.Space’s March 5, 2026 recap reported that he described Gemini 3.1 Flash-Lite as “incredibly fast and cost-efficient” for its performance.[2]

Before DeepMind and AlphaFold, Hassabis worked in game development and studied the brain. He designed and programmed Theme Park at age 17, subsequently graduated from Cambridge University, and completed a PhD in cognitive neuroscience at University College London. His curriculum vitae lists his doctoral study at UCL’s Institute of Neurology as 2005–2009. This background provides context for his later public arguments about the relationship between neuroscience and AI.[6][7][4]

In the August 2017 commentary “AI and Neuroscience: A virtuous circle”, Hassabis, Christopher Summerfield and Matt Botvinick argued that neuroscience could help validate existing AI techniques and inspire new algorithms and architectures. They also advocated exchange in the other direction: AI research could help explain biological brains.[4]

Their concrete example was experience replay in DeepMind’s deep-Q network, DQN. Biological brains replay earlier patterns of activity during rest; DQN stores a subset of training data and revisits it offline to learn from earlier successes and failures. The coauthors presented this mechanism as an example of neuroscience informing AI, and identified efficient learning, imagination and transfer learning as further opportunities for exchange.[4]