AlphaGo

AlphaGo is an artificial intelligence program developed by DeepMind (owned by Google/Alphabet) that plays the board game Go. It combines deep neural networks with Monte Carlo tree search to evaluate positions and select moves. In March 2016, AlphaGo defeated Lee Sedol, the world's top-ranked Go player, in a five-game match (4–1), marking a watershed moment in AI research.

Go is significantly more complex than chess—with roughly 10^170 possible positions compared to chess's 10^47—making it a long-standing grand challenge for AI. Traditional brute-force approaches were infeasible. AlphaGo's breakthrough lay in combining supervised learning (trained on professional games) with reinforcement learning (self-play improvement) and a value network (position evaluation) alongside its policy network (move selection).

The victory demonstrated that deep learning and neural networks could master intuitive, pattern-recognition tasks previously thought to require human creativity and judgment. AlphaGo's success catalyzed broader investment in AI research, influenced public perception of AI capabilities, and validated the deep reinforcement learning paradigm. Subsequent versions (AlphaGo Zero, AlphaZero) improved further by learning entirely from self-play without human game data.