Experiments with AI to play games Monte-carlo thingy Neural network things I'm trying to accomplish this by researching as little as possible. Some ideas from Metaswitch Engineering Conference talk.