Explaining explainable AI

What is explainable AI, and why is it such a hot topic today?