When we use media, our experiences are now curated by algorithms that act without providing an explanation of their process. Shows are recommended, ads are placed, products are priced, and some communications are highlighted while others are ignored, all without any clear sense of a designer or motive. In this talk, I consider cases where machine learning (ML) systems in this domain produce quixotic and unexpected results because of the tendency toward “promiscuous association.” It has often been useful to speak of technological systems as having designers, and indeed to say that a system was “designed” in the first place implies a motive behind its operation. In ML, I argue that making some decision-making algorithms more accountable by attributing a design, process or motive to them may never work. If this avenue is closed, how can we seek understanding of and accountability for these systems?