The tutorial connects LLM mechanisms to three music-facing roles: listener, producer, and performer. The goal is not merely to enumerate applications, but to understand which capability enables each workflow and how to verify it.
Open questions¶
How can models demonstrate that an answer is grounded in musical evidence?
Which evaluation protocols predict usefulness in real creative workflows?
How should systems balance general language ability with specialist music tools?
What interaction patterns preserve human agency and musical intent?
Q&A and discussion¶
TODO: Collect audience questions and links to follow-up material here.