Unveiling AI Complexity

Tim and Lancelot discuss the complexity of AI learning, highlighting how Alphago's self-play revealed a higher skill ceiling than training on human episodes. They delve into the balance between core knowledge and unconstrained learning, shedding light on the challenges of communication with highly intelligent agents.