The application of machine learning algorithms as applied to physics problems have been gaining traction. In this talk, I will cover two topics. The first is a study of a novel approach to organizing training data, so-called weak supervision. A weekly supervised neural network is trained using mixed data sets, and the fraction of signal is provided to the algorithm. I will discuss recent work that explains why this approach is useful, and will explore its robustness. Next I will provide a scheme for asking the question “What is the Machine Learning?”. This “data planing” approach is a procedure for identifying combinations of variables -- aided by physical intuition -- that can discriminate signal from background. Weights are introduced to smooth away the features in a given variable(s). New networks are then trained on this modified data. Observed decreases in sensitivity diagnose the variable's discriminating power. Planing also allows the investigation of the linear versus non-linear nature of the boundaries between signal and background. We demonstrate the efficacy of this approach using a toy example, followed by an application to an idealized heavy resonance scenario at the Large Hadron Collider. By unpacking the information being utilized by these algorithms, this method puts in context what it means for a machine to learn.