He recently did an interview with Lex Friedman (<a href="https://www.youtube.com/watch?v=P-2P3MSZrBM" rel="nofollow">https://www.youtube.com/watch?v=P-2P3MSZrBM</a>) and it was pretty good. It's a bit long but worthwhile. Covers a bunch of topics about the philosophy of machine learning, models of thinking and consciousness, and some bits and pieces of future directions for AI.