Posted on October 18, 2017 by New Scientist - NewsAlphaGo’s AI upgrade gets round the need for human input AlphaGo Zero, Google DeepMind's artificially intelligent Go player, dominates humans and other AIs by learning itself – without any human training