Generative AI is everywhere. Each campaign presents it as an efficiency fix, often for problems it invented. Yet efficiency collapses in fields that thrive on the very search for problems. In the creative arts, including music, the goal usually is not to solve problems; it’s to seek them, wrestle with them and forge novel, unexpected forms of expression. That clash between algorithmic speed, and the creative value of exploration sets the stage for this session.

In this lecture, composer and AI researcher Misagh Azimi will showcase headline systems created to generating music and discuss the politics and carbon costs embedded in their creation, training, and data use. We’ll then trace the economic cracks beneath copyright, royalty streams, and cultural value, before outlining ways to intervene through personal fine-tuning, transparent datasets, and collaborative, human-centered workflows.