The host discusses the potential for making new expositional contributions to the cutting edge science of machine learning, particularly through multidimensional visualizations, even if one does not have the background to make new scientific discoveries themselves.