Description
Over the past few years, generative audio models have moved from emerging technologies to increasingly established tools for musical creation. As artists, researchers, and developers have gained experience working with diffusion models, neural audio synthesis, latent-space instruments, and other generative approaches, it has become possible to look beyond their initial promise and ask what these technologies actually afford in practice.
This new edition of the Generative Audio for Music-Making (GAMMa) workshop invites the community to reflect critically and creatively on what we have learned from making music with generative audio systems. What new forms of musical interaction, expression, and experimentation have they enabled? How have they influenced our approaches to sound, instrument design, composition, improvisation, and performance? What happens when these models are coupled with bodies, sensors, and physical instruments, or trained on small, bespoke datasets? Where do current models fall short, and what technical, artistic, ethical, or practical limitations have surfaced through their use? What would we like the next generation of these technologies to do differently?
Call for Participation
The workshop welcomes a broad range of perspectives, including theoretical, practical, analytical, and inquisitive presentations. Contributions may address artistic practices, technological developments, critical perspectives, case studies, reflections on creative processes, or emerging research questions. Demonstrations of instruments, systems, software, and works-in-progress are also encouraged.
If you would like to submit a presentation along those lines, please fill out the following form before Nov 19th 23:59 ET: https://forms.cloud.microsoft/r/jaLCDdNtkh
GAMMa 2026 aims to create a space for sharing experiences—not only successes, but also limitations, unexpected outcomes, productive failures, and unresolved questions—and for collectively considering where generative audio for music-making might go next.