Boon, H. 2026. Music and AI Problematics: Interface Design and the Current Lack of Musical Correlates. 9th Annual School of Computer Science & Engineering Student–Staff Research Conference. University of Westminster 25 Feb 2026
| Title | Music and AI Problematics: Interface Design and the Current Lack of Musical Correlates |
|---|---|
| Authors | Boon, H. |
| Abstract | This presentation concerns the disconnect between current AI interface designs and what H. Stith Bennett referred to as the historical framework of “music’s technical features” (Bennett, 1983). Arguably, AI and its current interfaces replace the traditional dominance of the piano keyboard, the long-influential history identified as Claviocentrism (Diduck, 2018), with simple prompts and a limited number of sliders, providing some user agency. However, the underlying processes and procedures used to produce the final output remain largely opaque or “Black Boxed” (Seaver, 2019). In terms of Black Boxing, Bringsjord highlights that, for those working in AI, the machine removes the “burden of having to express a problem and a solution to a problem in precise terms” (Bringsjord, 2022). This he refers to as a dead end, which produces the magic. Rather than notation and embodied musical performance as the primary historical technical features, AI systems introduce new technical elements that rely primarily on recorded work, which has a shorter history (Boon, 2022a; Boon, 2023). This shift unseats the central historical features that many composers and musicians previously used to craft their identities and establish their reputations (Boon 2024). In sum, a shift from embodied to disembodied that, rather than introducing new forms, historicises past periods. This presentation identifies the problem of current AI interface design as the lack of musical correlates and musical fit (Deruty et al., 2022; Huang et al., 2020; Tchemeube et al., 2021; Valenzuela, 2021). As music AI systems require the ingestion of recorded material, this necessitates developing business relationships with small but powerful groups of rights holders as an act of “accumulation by dispossession” (Boon, 2022b) rather than establishing direct connections with musicians, as was common in the previous system. This absence of musical features in AI applications poses a challenge, perhaps suggesting a need for AI interface design to bifurcate into consumer and musician/producer systems. |
| Keywords | Music Production |
| Artificial Intelligence | |
| Generative AI | |
| Interface design | |
| Black box | |
| Digital Alienation | |
| Accumulation by dispossession | |
| Year | 2026 |
| Conference | 9th Annual School of Computer Science & Engineering Student–Staff Research Conference |
| Publication dates | |
| Published | 25 Feb 2026 |
| File | License CC BY-SA File Access Level Open (open metadata and files) |