Music poses unique and complex challenges for artificial intelligence, even as 21st-century AI grows ever more adept at generating compelling content. The AI Music Problem: Why Machine Learning Conflicts With Musical Creativity probes the challenges behind AI-generated music, with an investigation that straddles the technical, the musical, and the aesthetic.

Bringing together the perspectives of the humanities and computer science, the author shows how the difficulties that music poses for AI connect to larger questions about music, artistic expression, and the increasing ubiquity of artificial intelligence. Taking a wide view of the current landscape of machine learning and Large Language Models, The AI Music Problem offers a resource for students, researchers, and the public to understand the broader issues surrounding musical AI on both technical and artistic levels.

The author breaks down music theory and computer science concepts with clear and accessible explanations, synthesizing the technical with more holistic and human-centric analyses. Enabling readers of all backgrounds to understand how contemporary AI models work and why music is often a mismatch for those processes, this book is relevant to all those engaging with the intersection between AI and musical creativity today.

ACKNOWLEDGMENTS

Many thanks to Jesse Caputo, Sarah Connors, Ian Quinn, Heather Peterson, Nicole Cosme-Clifford, Kavi Kapoor, James Symons, Leo von Mutius, Elizabeth Medina-Gray, Megan Long, Chris Brody, Scott White, Andrew Goldman, Catrina Kim, Mariusz Kozak, Alex Rehding, Justin London, Mark Gotham, Dan Harrison, and my terrifyingly brilliant husband Robert Powell for their insights and input into this project.