AlphaZero

About

• AI system developed by DeepMind
• Introduced in 2017 as a generalization of AlphaGo Zero
• Learned entirely through self-play and reinforcement learning
• Mastered Go, Chess, and Shogi
• Did not require human gameplay data
• Uses deep learning with Monte Carlo Tree Search
• Influential example of general-purpose game-playing AI

See also

• AlphaGo | AlphaGo Zero
• Google DeepMind | Reinforcement Learning