Machine learning is now ubiquitous in modern society. It's often used to solve problems such as spam email detection, natural language translation, to find images matching certain search criteria, or to make music recommendations. This talk will turn the problem around and examine the extent to which neural networks can emulate human creativity. We'll start by thinking about what it means for an algorithmic system to be creative. Then through examples we'll look at how techniques in machine learning can be used to compose original music and generate art. During the talk, we'll see how changing the algorithms and modifying the inputs to the neural networks affect the music and art created. We'll conclude by discussing future directions in the field of computational creativity, and thinking about whether computers will ever be able to match human levels of performance in this area.