There is DeepChess, which is a proof of concept that it can be done: https://www.cs.tau.ac.il/~wolf/papers/deepchess.pdf https://www.cs.tau.ac.il/~wolf/papers/d

There is DeepChess, which is a proof of concept that it can be done: https://www.cs.tau.ac.il/~wolf/papers/deepchess.pdf https://www.cs.tau.ac.il/~wolf/papers/deepchess.pdf I don't think it will necessarily be as good at chess, because chess has a much smaller search space. So "stupider" methods can actually work, while Go requires very good pattern recognition to prune the search space.