DeepMind Workshop

Alex Davies, University of Cambridge

What is Machine Learning Good For?

Machine learning has seen remarkable success in many areas in the past decade, but this inevitably leads to hype that obscures what it can and cannot do. In this talk we will address this by covering how machine learning problems are generally framed, what properties of problems makes them hard/easy to address with machine learning and examples of how machine learning was used to make advances in different fields. These examples will include recent work such as our first publications in mathematics, AlphaGo and AlphaFold. This will provide an overview that requires no prior experience with machine learning, but will directly address the current frontier of machine learning.