John Schulman - The Future of Reinforcement Learning (via searchlore.ai)