We can build and train a neural network using any large data set, and then use it to discover key features in the data. For example, using the Titanic passenger list, which includes data on survival, we can identify characteristics that seem to lead to some passengers surviving. This is a Humanities-oriented application of the technology. We might also use similar techniques creatively, for example, generating virtual worlds populated with people similar to those on the Titanic. In this episode we explore neural networks and techniques for using them in the Arts and Humanities. It includes a good enough explanation of how they actually work, without going into too much technical detail.