In this talk, Dr David Hall will give a high-level overview of supervised deep learning and how it may be applied to applications in weather, climate, and the earth system sciences. Supervised deep learning is a new and powerful technique for building software from data. In contrast with traditional programming, deep learning functions are built automatically from a labelled dataset, describing what you want the function to do, rather than how to do it. With this approach, we are able to build next-generation software capabilities that were previously unachievable. Some promising applications of AI for the Earth system include automated feature detection and data thinning, strategic planning, GPU model acceleration, image enhancement, more accurate physical parameterizations, and more. As with any technique, deep learning comes with its own unique set of challenges, which David Hall will discuss, as well as techniques being developed to overcome them.