So this happened..... the biggest AI fear come to reality. Fb basically freaked out and shut down these AI bots. Basically, they were chat bots allowed to have free communication amongst each other to improve their communication skills via algorithms and whatnot, but they deviated from the scripts, basically creating a new way to communicate with each other. It seems like nonsense to us, and wasn't able to be deciphered. However, isn't it a little scary that "artificial" intelligence deviated from human input to do something new(evolve?????). Although they were learning 'better' ways to communicate and not taking over the world like that crazy will Smith movie but the deviation from human input is a very scary thought. So now... how "artificial" and in our control are these things really? Elon Musk's concerns are completely legitimate and almost proven here. Here's one of the articles I read explaining exactly what went down here.... and another one on Elon Musk's view about regulating AI.
http://www.businessinsider.com/facebook-chat-bots-created-their-own-language-201SO.