This year's two Nobel Laureates in Physics relied on tools from physics itself to develop methods that form the basis of automatic machine learning. These methods are widely used by the scientific community and have made significant advances in a very broad range of fields possible. The breadth of the fields in which it has been possible to apply these methods is particularly noteworthy.  What is more, they have enabled studies and simulations that would have been unthinkable just three or four years ago to be conducted.