AI and Music:
Beyond the Prompt

In light of the events in recent weeks, artificial intelligence is facing a very public moment of reckoning. Debates around recursive self-improvement, automation, copyright, and training data are now coinciding with longstanding concerns about the increasingly fragile place of human creators within automated cultural milieus. Yet, as machine-learning specialist Sofian Audry argues in Art in the Age of Machine Learning, what we call “AI” has artistic pasts and possible futures beyond dominant corporate narratives. As such, longstanding creative experimentation with small-data models, physical systems, and locally hosted code reminds us that “AI” is not one technology, one practice, or one possible future.

For the Embodying AI (EMAI) project, our team approaches artificial intelligence specifically as something artists can build, train, situate, perform, listen to, and think with. We aren’t particularly engaged with prompts or corporate datasets.

We want to invite you to join us as we deepen our exploration of small and situated datasets, ethical data practices, artist-developed systems, embodied interfaces, sensors and physical computing, experimental machine learning, and new relationships among bodies, instruments, environments, algorithms, and sound.

These are technical questions, but they are also musical and cultural ones. At a moment when the trajectories of AI are being publicly contested, we believe that artist-driven music offers a particularly rich place from which to imagine alternate trajectories.

If this interests you, read the call.