It goes the other way too. Reinforcement learning and control theory are also very useful for general machine learning. That's part of the problem - engineers,

It goes the other way too. Reinforcement learning and control theory are also very useful for general machine learning. That's part of the problem - engineers, statisticians and computer scientists keep inventing different views of the same things. Now that people are communicating and consolidating things will start to get even more interesting.