AlphaGo

AlphaGo is a game-playing AI known for using search ideas to play the board game Go.

What it does

AlphaGo is a game-playing AI known for using search ideas to play the board game Go.

How it works

• Look at possible moves.
• Search ahead to imagine what may happen.
• Choose a strong move.

Uses and applications

• It plays the game Go at a very high level.
• Researchers use it to study smart search.
• Game makers can learn how AI compares future moves.

In short

AlphaGo is a well-known example of tree search used to choose moves.

Related methods

• MCTS (Tree Search): The tree search at the core of AlphaGo.
• AlphaZero: Learned from self-play alone, without human games.

Back to Reinforcement learning, agents and planning · All topics

The interactive demonstration and the Python walkthrough for this topic need JavaScript.