I would say that reinforcement learning is the same principle that humans learn and so it is in use, just at an early stage of development. The hard part is specifying the goal (or sequence of goals) correctly in a way a computer can understand.

Source: Artificial Intelligence on Medium