Machine learning is changing our ability to recognize patterns in the world, act on them, and generate change. The algorithms have typically been implemented with the goal of "replacing" error-prone human hands. By performing tasks better than humans, costs can be minimized and success can be increased. Importantly, this process implies that machines have become better and better at encoding information about the world that human individuals have learned to identify through experience, and they are improving on these capacities. But what about the possibility of machines learning what collectives of humans know how to do? Can machines learn how countries develop, and if they can, can they accelerate the process? Informed by recent developments in the field of economic complexity, we explore machine learning methodologies to predict diversification events in the export baskets of countries.