Thore Graepel
Thore Graepel is a machine learning researcher who worked on systems such as TrueSkill at MicrosoftMicrosoftMicrosoft is a technology company founded by Bill Gates and Paul Allen in 1975 and headquartered in Redmond, Washington, with businesses spanning Windows, Microsoft 365, Azure, enterprise software, artificial intelligence, gaming and devices.Open the full entry Research and later contributed to AlphaGo at DeepMindGoogle DeepMindGoogle DeepMind is Google’s artificial intelligence research organization, formed in April 2023 by merging DeepMind, which Google acquired in 2014, with the Brain team from Google Research, and led by CEO Demis Hassabis. It developed AlphaGo, AlphaFold, and the Gemini family of models, and Hassabis and John Jumper received the 2024 Nobel Prize in Chemistry for AlphaFold’s protein structure prediction work.Open the full entry; he is also Chair of Machine Learning at University College London and returned to Google DeepMind in 2025.
Contents10 sections
Key facts
2001PhD from the Technical University of Berlin
He studied physics at the University of Hamburg, Imperial College London and the Technical University of Berlin, where he received a PhD in machine learning in 2001; he then held postdoctoral positions at ETH Zurich and Royal Holloway College, University of London.[1]
2003Joining Microsoft Research Cambridge
In 2003 he joined Microsoft Research in Cambridge, where he co-founded the Online Services and Advertising group; his work was applied in Xbox Live’s TrueSkill ranking and matchmaking system, the AdPredictor click-through-rate prediction framework in Bing, and the Matchbox recommender system, which inspired the recommendation engine of Xbox Live Marketplace.[1]
March 2013Co-Authoring a Study on Predicting Private Traits from Digital Records
On March 11, 2013, he, Michal Kosinski and David Stillwell published online in the Proceedings of the National Academy of Sciences (PNAS) a paper on how private traits and attributes can be predicted from digital records of human behavior. The Cooperative AI Foundation said this work made him well known in the data privacy field.[2][3]
January 2016Co-Authoring the AlphaGo Paper
On January 27, 2016, Nature published online the DeepMind paper “Mastering the game of Go with deep neural networks and tree search,” of which he was a co-author. The Partnership on AI said that at DeepMind he contributed to creating AlphaGo, the first computer program to defeat a human professional player in the full-sized game of Go.[4][1]
October 2017Co-Authoring the AlphaGo Zero Paper
On October 18, 2017, he was among the authors of the paper “Mastering the game of Go without human knowledge,” published in Nature.[5]
February 26, 2020Joining the Board of the Partnership on AI
On February 26, 2020, the Partnership on AI announced that he was joining its board, taking over Mustafa Suleyman’s DeepMind seat; at the time he was a research group lead at DeepMind and held a part-time position as Chair of Machine Learning at University College London.[1]
March 2024At Altos Labs
According to a March 2024 speaker announcement by Air Street Press, he was then global head of computational science, AI and machine learning at the biotechnology company Altos Labs.[6]
October 2025Returning to Google DeepMind and Joining the Cooperative AI Foundation’s Board
On October 30, 2025, the Cooperative AI Foundation announced that he had joined its board of trustees, saying he had recently returned to Google DeepMind to lead a team working on post-AGI futures; the foundation also described him as an author of two papers that launched cooperative AI as a field and as a widely published researcher on multi-agent learning.[3]
Related Organizations, Websites, and Public Accounts
Google DeepMind: Official website: https://deepmind.google/ (opens in a new window)[7]
Partnership on AI: Official website: https://partnershiponai.org/ (opens in a new window)[1]
Cooperative AI Foundation: Official website: https://www.cooperativeai.com/ (opens in a new window)[3]
Sources
- Welcoming New PAI Board Members: Rebecca Finlay, Prof. Thore Graepel, and Dr. Prem Natarajan (opens in a new window)
- Private traits and attributes are predictable from digital records of human behavior (opens in a new window)
- Thore Graepel Joins Board of Trustees (opens in a new window)
- Mastering the game of Go with deep neural networks and tree search (opens in a new window)
- Mastering the game of Go without human knowledge (opens in a new window)
- RAAIS 2024 Speaker: Thore Graepel (opens in a new window)
- About Google DeepMind (opens in a new window)