But at some point Advanced Chess will be 'dead' in the sense that the best human players no longer help the best software win: the human will pick better moves less than poorer moves, and blow games. It's hard to tell when, but given how large the margin is, and how finely balanced the best chess engines are now (Lipton has some interesting posts on computer chess, most recently https://rjlipton.wordpress.com/2014/12/28/the-new-chess-worl... ), I wonder if that moment has already passed?
Another observation is that humans only add value for their pattern matching ability. Computers have worse heuristics, but make up for it by being able to explore many moves into the future.
But now deep neural networks have been catching on. They've shown good results on Go, being able to predict the move an expert would make 44% of the time.
