After all kinds of chess can defeat humans, DeepMind has finally turned its attention to those who beat the human race.
Yesterday, the AlphaGo research team proposed AlphaZero: A new algorithm that can reach beyond humanity on many tasks from scratch through self-reinforcing game learning.
That AlphaZero how powerful it:
8 hours training to beat Li Shishi version of AlphaGo
12-hour training to beat the world's top chess program Stockfish
14-hour training to beat the world's top chess program Elmo
Even in the face of the powerful AlphaGo Zero, AlphaZero also beat the former 72 hours of training after 34 hours of training, and careful friends might have noticed that the two were actually a 'Go' because of DeepMind's latest AlphaZero has gone beyond Go's boundaries, so removed as a representative of 'Go'.
Faced with this kind of all kinds of chess system AI Xiaobian just want to say, after you have children, do not train in the direction of professional players it.
After all kinds of chess can defeat humans, DeepMind has finally turned its attention to those who beat the human race.
Yesterday, the AlphaGo research team proposed AlphaZero: A new algorithm that can reach beyond humanity on many tasks from scratch through self-reinforcing game learning.
That AlphaZero how powerful it:
8 hours training to beat Li Shishi version of AlphaGo
12-hour training to beat the world's top chess program Stockfish
14-hour training to beat the world's top chess program Elmo
Even in the face of the powerful AlphaGo Zero, AlphaZero also beat the former 72 hours of training after 34 hours of training, and careful friends might have noticed that the two were actually a 'Go' because of DeepMind's latest AlphaZero has gone beyond Go's boundaries, so removed as a representative of 'Go'.
Faced with this kind of all kinds of chess system AI Xiaobian just want to say, after you have children, do not train in the direction of professional players it.
