David Silver is a British researcher and university professor, recognized worldwide as one of the leading authorities in the field of deep reinforcement learning.
His work at Google DeepMind completely transformed the landscape of artificial intelligence, demonstrating that algorithms can learn master strategies through direct interaction with their environment, surpassing limits that scientists believed were unattainable for decades.
Academic training and career at DeepMind
Silver graduated in Computer Science from the University of Cambridge and completed his PhD at the University of Alberta under the supervision of Rich Sutton, considered the father of reinforcement learning. In 2013 he formally joined DeepMind Technologies, where he assumed leadership of the research team on reinforcement learning algorithms.
The historic revolution: From AlphaGo to AlphaZero
The big turning point in his career occurred in March 2016 with the development of AlphaGo. The system made global history by defeating the legendary champion Lee Sedol 4-1 in the ancient game of Go, a challenge that the scientific community considered impossible to solve in the short term due to its combinatorial complexity.
Subsequently, Silver led the creation of AlphaZero, an even more momentous advance: a unique algorithm capable of learning from scratch (through self-play and without prior knowledge provided by humans) to master chess, shogi and Go itself in a matter of hours, demonstrating the viability of generalized reinforcement learning.
Global impact and applications beyond games
The techniques developed by David Silver are not limited to board games. His research in policy networks and value networks is applied today in frontier fields such as autonomous robotics, energy optimization in data centers, the discovery of new drugs and advanced materials science.
Explore more profiles in our directory: Meet all the scientists, founders and leaders who are setting the course for technology in our pillar guide: Who’s who in AI?.
Cet article est également disponible en :