Self play

3 entries between August 2017 and December 2017, 1 turning point.

Turning points

  1. AlphaGo Zero learns Go from self-play alone

    DeepMind announced AlphaGo Zero, trained only by playing itself from random play, with no human games and no hand-crafted features. After three days it beat the version that had defeated Lee Sedol, 100 games to nil. Nature published the work the next day.

Every entry

  1. AlphaZero learns chess and shogi from the rules alone

    DeepMind posted a paper describing AlphaZero, one algorithm that, given only the rules, reached superhuman play at chess, shogi and Go within 24 hours of self-play training and beat the reigning computer champion program in each game.

  2. AlphaGo Zero learns Go from self-play alone

    DeepMind announced AlphaGo Zero, trained only by playing itself from random play, with no human games and no hand-crafted features. After three days it beat the version that had defeated Lee Sedol, 100 games to nil. Nature published the work the next day.

  3. An OpenAI bot beats a professional Dota 2 player

    A bot trained by OpenAI beat the Ukrainian professional Danil Ishutin, known as Dendi, in a one-on-one exhibition match at The International. OpenAI said it had learned by playing itself, gaining experience equal to about two weeks of continuous play.