Richard Sutton

Sutton co-authored 'Reinforcement Learning: An Introduction' with Andrew Barto, first published by MIT Press in 1998 with a second edition in 2018. The book is freely available online and covers Markov decision processes, temporal-difference learning, and policy gradient methods, making it the standard reference cited tens of thousands of times.

Sutton developed temporal-difference (TD) learning, a core reinforcement learning method that combines Monte Carlo sampling with dynamic programming bootstrapping to learn from incomplete episodes of experience. His 1988 paper on TD learning is among the most cited works in the field and underpins algorithms like TD-Gammon and modern deep RL.

Sutton's research directly underpins DeepMind's AlphaGo, which defeated world champion Lee Sedol at Go in 2016, and the reinforcement learning from human feedback (RLHF) techniques used to train large language models like OpenAI's GPT series. His emphasis on learning through interaction with an environment remains central to AI research toward general intelligence.

Sutton is a professor at the University of Alberta in Edmonton, Canada, where he has been since 2003. He also served as a distinguished research scientist at Google DeepMind's Edmonton office from 2017 until 2023, when he returned to full-time academic work.