The Future of Reinforcement Learning

Julian Tugelius: I'm definitely very much into this sort of level generators and game-playing or acting in an environment agents. There's a lot of interesting stuff to work in on there. Some people got it open ended learning, which is a good term for it, I guess. But you can basically use reinforcement learning in every stage - not saying that reinforcement learning is necessarily the best way of doing this. He says his heart still beats for evolutionary algorithms.