A Context-Sensitive Approach to XAI in Music Performance

One-size-fits-all explanations don't work for AI in creative fields. The paper argues that there is "no single approach that can be universally applied" to explain AI systems, particularly in the arts like music performance, challenging the assumption that a standard explainability method can serve all purposes.

Context and audience matter more than technical complexity. The Explanatory Pragmatism framework emphasizes that effective AI explanations must be tailored to who is asking and why they're asking, rather than providing generic technical breakdowns—meaning explanations for musicians should differ from those for audiences or developers.

Explanations should be iterative and responsive. The paper proposes that explainability requirements should be "continuously refined based on feedback," suggesting that responsible AI implementation in creative domains requires ongoing dialogue with stakeholders rather than one-time documentation.

Artistic AI demands a different approach to transparency. The paper identifies the arts as a domain where standard XAI methods are especially inadequate, implying that responsible AI governance needs to account for different industries and creative applications, not just high-stakes domains like healthcare or criminal justice.