Machine learning presents some tough questions when applied to investing. What do you do when a model, which has always worked well in the past, all of a sudden predicts something that seems crazy, but you don’t know why? Is it insane… or genius? And how would you explain your decision it to your own clients or investors? Machines have a lot to say, if we can figure out how to listen. Andrew Li, Quantitative Researcher at State Street Associates, discussed key principles of machine learning, model interpretability frameworks, and how machines and humans can help each other improve.