3 ms·
From the paper: > Defeat the best Go players, training only on as many games as the best Go players have played. For reference, DeepMind’s AlphaGo has probabl
by throwawaymsft 9y ago
From the paper:
> Defeat the best Go players, training only on as many games as the best Go players have played. For reference, DeepMind’s AlphaGo has probably played a hundred million games of self-play, while Lee Sedol has probably played 50,000 games in his life.
- 6nf 9y agoSeems a bit unfair. Go player study games of other players too, they don't just rely on personal experience.
- majewsky 9y agoBut it's still much less than 100 million games. If I assume that playing a Go game (or looking at a record and understanding the moves taken) takes one hour (which is a wild guess since I'm not familiar with the game), 100 million games take nearly 11408 years.