After implementing a very basic hide-and-seek game, and teaching the artificial intelligence players simple rules, reinforcement learning led to truly fascinating results. Have a look at the animated gifs in the article, and remember that none of these specific tactics were programmed by humans.