Achieving provably beneficial, human-compatible AI (Practical AI #85)

AI legend Stuart Russell, the Berkeley professor who leads the *Center for Human-Compatible AI*, joins Chris to share his insights into the future of artificial intelligence. Stuart is the author of *Human Compatible*, and the upcoming 4th edition of his perennial classic *Artificial Intelligence: A Modern Approach*, which is widely regarded as the standard text on AI. After exposing the shortcomings inherent in deep learning, Stuart goes on to propose a new practitioner approach to creating AI that avoids harmful unintended consequences, and offers a path forward towards a future in which humans can safely rely of provably beneficial AI.