"In our view reinforcement learning is going to be as big as deep learning in the next two or three years." - Demis Hassabis in Technology Review portrait • /r/MachineLearning