AlphaGo

AlphaGo is the DeepMind system that won four of five games against Lee Sedol in March 2016. Go had long seemed out of reach for brute-force search alone because the board is vast and expert play feels intuitive. It followed IBM’s Deep Blue chess milestone (1997), where search without learning had already beaten a world champion.

AlphaGo combined neural networks, reinforcement learning from self-play, and Monte Carlo tree search. It became a public symbol of deep learning moving from pattern recognition into long-horizon strategy, not the same as a chat LLM, but part of the same modern ML toolkit.