Currently obsessed with reinforcement learning. What's next?