Differential and Probabilistic Programming for Scientific Computation

In this talk, I will discuss where I see opportunities to apply AI in scientific computation. Whilst the mainstream of work in AI has predominantly focused on images and text, there lies a tremendous potential for AI in domains where data modalities and prediction tasks are much more diverse. I will give examples of applications of deep learning, differential programming, and probabilistic programming to these domains, and conclude with a perspective on how these methods can be combined to overcome bottlenecks in scientific computation.