Loading timeline…
20162010s
AlphaGo / AlphaZero
The AI system that conquered the ancient, mathematically intractable game of Go.
Why It Was Important
DeepMind's AlphaGo defeated world champion Lee Sedol, a feat experts believed was decades away due to Go's impossibly vast search tree. AlphaZero later learned Chess, Shogi, and Go purely from self-play (tabula rasa), discovering entirely new, alien strategies that revolutionized how humans play the games.
Who Invented It
David Silver, Demis Hassabis (DeepMind)
The AGI-focused team merging deep neural networks with Monte Carlo Tree Search.
Applications
- Board Game Engines
- Protein Folding (AlphaFold precursors)
- Reinforcement Learning
Key Papers
- Mastering the Game of Go with Deep Neural Networks and Tree Search
David Silver · Nature · 2016
- A General Reinforcement Learning Algorithm that Masters Chess, Shogi, and Go through Self-Play
David Silver · Science · 2018
Videos
AlphaZero: Shedding new light on the grand games of chess, shogi and Go
Google DeepMind