Researchers behind AlphaGo and the planning systems that followed it.
20 people
Google DeepMind +2
Chair
Hassabis helped turn AI into a tool for scientific discovery. He co-founded DeepMind and led the teams behind AlphaGo and AlphaFold, sharing the 2024 Nobel Prize in Chemistry for work predicting the three-dimensional shapes of proteins.
Ineffable Intelligence +1
Founder and CEO
Silver led AlphaGo, which defeated a world champion at Go, and AlphaZero, which learned chess, shogi, and Go largely through self-play. These systems showed that reinforcement learning could discover powerful strategies from experience, an approach he is now extending beyond games through Ineffable Intelligence.
Metis Reasoning +1
CEO & Co-Founder
Graepel helped create AlphaGo and co-authored AlphaGo Zero, AlphaZero, and MuZero—systems that showed how an AI could learn difficult games through practice and planning rather than human examples. He also created widely used methods for matchmaking and online advertising.
Periodic Labs
Guez helped create AlphaGo and co-authored AlphaGo Zero, AlphaZero, and MuZero. These systems showed that an AI could combine learning and look-ahead planning to master difficult games, with MuZero doing so without being given the game's rules. His work helped establish planning as a core capability of learned agents.
Anthropic
Member of Technical Staff
Schrittwieser helped build AlphaGo, AlphaZero and MuZero, systems that showed how neural networks and planning could master complex games with progressively less human guidance. He later contributed to AlphaCode, AlphaTensor, Gemini and AlphaProof.
Google DeepMind
Research Engineer
Hubert helped build DeepMind's line of self-play and planning systems: AlphaZero showed one reinforcement-learning algorithm could master chess, shogi, and Go from the rules alone; MuZero pushed the idea further by planning with a learned model of the environment.
Google DeepMind
Senior Staff Research Scientist
Huang was the lead engineer behind AlphaGo—the system that beat the world's best Go players—and physically placed its stones in the historic matches. He is a researcher at Google DeepMind.
Reflection AI
Co-founder, President and CTO
Antonoglou contributed to the deep Q-network that learned Atari games from pixels, then to AlphaGo, AlphaZero, and MuZero. Across these systems, he helped show that one neural architecture could learn perception, planning, and strategy from experience rather than relying mainly on hand-written rules.
H Company
Co-founder and CTO
Sifre contributed to AlphaGo and developed a more efficient way for vision models to process images. That design sharply reduced the computing needed for image recognition and became a core building block in widely used systems such as MobileNet and Xception.
Persistence Bio +2
Co-Founder
Maddison helped build AlphaGo, the first program to defeat a world champion at Go. He developed widely used methods for estimating how to improve neural networks when they make distinct choices. He also applies machine learning to drug discovery, where useful data is scarce.
Google DeepMind
Research Engineer
Van den Driessche co-authored the original AlphaGo and AlphaGo Zero Nature papers, then contributed to language-model systems including RETRO, Chinchilla, Gemini, and Gemini 2.5.
Phaidra
CTO and Co-Founder
Panneershelvam—credited as Veda Panneershelvam on the 2016 Nature paper—was a key engineer who designed and built AlphaGo. At Phaidra, he leads AI research and engineering for autonomous control of data centers and other industrial systems.
Google DeepMind
Research Scientist
Lanctot develops methods for agents that learn in competitive and cooperative games. He helped create AlphaGo and founded OpenSpiel, an open framework that lets researchers compare reinforcement-learning, planning and game-theoretic algorithms across many kinds of games.
Google DeepMind
Research Scientist
Dieleman has helped shape modern AI across games and generative media. He co-authored the AlphaGo paper and contributed to WaveNet, Lyria, Imagen, and Veo. His research focuses on diffusion models, which create media by repeatedly refining noise, and on learning useful structure in sound, images, and video.
Isomorphic Labs
Principal Software Engineer
Grewe co-authored DeepMind's 2016 AlphaGo paper and spent more than a decade building tools and libraries that accelerated AI research. His doctoral work focused on compiling and optimizing parallel programs for heterogeneous hardware; he now builds software for AI-driven drug discovery at Isomorphic Labs.
OpenAI
Research
Nham helped build AlphaGo, the landmark system that defeated a world champion at the game of Go. He later worked on making language models more factual and now develops systems that learn to reason through difficult problems.
Project Prometheus
Founding Member
Kalchbrenner helped establish neural generation across several kinds of data: PixelCNN for images, WaveNet for high-fidelity speech, and ByteNet for parallel sequence processing. He also contributed to AlphaGo and neural weather forecasting, showing how related learning ideas could transfer across very different domains.
Safe Superintelligence Inc.
Co-founder and CEO
Sutskever co-created AlexNet, a breakthrough in image recognition, and sequence-to-sequence learning for translating sentences. As OpenAI’s co-founder and chief scientist, he helped shape the GPT models and training larger models that led to ChatGPT. In 2024 he left to co-found Safe Superintelligence, aiming to build AI smarter than humans safely.
UCL +1
Adjunct Professor
Lillicrap co-developed DDPG for continuous control and contributed to AlphaGo and AlphaZero at Google DeepMind, working at the intersection of deep reinforcement learning and neuroscience.
Google DeepMind +1
Senior Vice President
Kavukcuoglu helped build DeepMind's deep-learning program and co-created DQN, which learned to play Atari games from raw screen images, and WaveNet, which made computer-generated speech sound more natural.