This seminar will work towards settling on a set of research questions about an inherently unsettled technological domain: AI/machine learning. Articulating and exploring research questions will not however exclude the practical and in fact much of the work we engage will either emerge from practice or be grounded in empirical inquiry. To attempt something like comprehensiveness for this field is implausible and indeed unfeasible, but we should be able to arrive collectively at the point of something like foundational understanding, with a clear sense of directions for future research and teaching. Although we will not proceed in this order, we will take an end-to-end approach, from data, through model architectures and the production of prediction, and ending with applications and implementations.