Sergey Levine - Self-Play in Reinforcement Learning (via searchlore.ai)