WP4a: ‘Localising’ Recommendation: Serving Middle Eastern Listeners and Music

Darci Sprengel, Lecturer, Department of Culture, Media and Creative Industries, King’s College London

This study relocates MusAI in the global South, addressing how recommendation systems tend to build their algorithmic infrastructures on reified Western musical epistemologies and ontologies (Born 2019, Gómez et al 2013). Through comparative ethnographic research on the AI infrastructures of Spotify’s regional operations and Anghami, two platforms serving the Middle East, the project investigates recommendation’s mediation of the listening subject in relation to locality and cultural difference. In so doing, it brings into focus how AI infrastructures may enact ‘algorithmic oppression’ (Noble 2018, Benjamin 2019), or – as in Anghami’s case – may attempt to model listening differently. The study brings feminist and critical race approaches to bear on ethnographic research among listeners and in both platforms’ Dubai offices, prototyping a collaborative model in which ethnomusicology can assist in developing more inclusive streaming services.