Spotify prompt-driven AI playlists risk devaluing many artists, genres

Spotify is testing AI-generated playlists based on custom prompts. The streamer successfully launched its AI-powered DJ feature last year.

Code pointing to prompt-driven AI playlists was discovered within the Spotify app by tech investor Chris Messina, who posted screenshots.

Prompt-driven AI creates – in this case, playlists – based on a few words proved by the user. Examples might be “upbeat K-pop” or “acoustic rock hits.”

One of Spotify’s strengths has always been the ability of its algorithms to mix new music discoveries with familiar tracks. I’ve heard many artists for the first time on a Spotify playlist that is now in regular rotation.

But it’s unclear whether prompt-based AI will also surface songs, artists, and even genres that are multifaceted and harder to describe. Or will it naturally only deliver tracks based on the common sonic vocabulary of the average fan?

How will this affect the many artists whose music does not fit neatly in a category or whose songs are more difficult to describe?

Bruce Houghton is the Founder and Editor of Hypebot, a Senior Advisor at Bandsintown, President of the Skyline Artists Agency, and a Berklee College Of Music professor.