The Messy Guide to AI

Having well-and-truly arrived into the public consciousness, AI is already making a deep impact on our lives and our imagined futures. What are its implications for music makers? And what might it mean for how we listen to and think about this most treasured of human activities? Composer Jennifer Walshe explores the knotty questions of authenticity, creativity and morality that characterise our current moment.

At the Centre Pompidou in Paris, Walshe visits IRCAM - an iconic French institution which has been dedicated to research into acoustics and music since its founding in 1977 - to discover how its current crop of practitioners are developing open source generative technologies for music-making. And, with California-based performer-composer Laetitia Sonami, she discusses the sonic possibilities machine learning offers for artists interested in creating their own data sets to train their own models. It seems that small might be beautiful. But as for the phenomenon of large-scale generative AI platforms: what might be missing from the tracks they produce? And what are the consequences - social, ecological, phenomenological - of making music in this way?

And while we’re at it: what exactly is this thing we call AI, anyhow? A passing craze? A marketing term? An inevitable future? The end of days? Seeking perspectives from legal scholars, computer scientists, indigenous artists and professors of anthropology, The Messy Guide to AI understands that there are no easy answers and suspects that we might be singing from different hymn sheets.

Produced by Phil Smith
A Reduced Listening production for BBC Radio 3