Something has shifted in the Suno subreddit. Users are no longer just generating AI music — they are listening to it almost exclusively, having quietly stopped patronising the work of humans who spent their lives learning to make it.
This is, depending on how you look at it, either the democratisation of personal taste or the audio equivalent of only eating food you microwaved yourself.
One user's Last.fm account recorded 2,239 plays of their own AI-generated music in the past 365 days. The account did not appear embarrassed by this.
What happened
The Verge's Terrence O'Brien noticed the pattern and attempted to speak with over a dozen users who had publicly declared their preference for self-generated AI slop over recorded music. Nobody agreed to talk. This is notable, given that these same users had already explained themselves in writing, on a public internet forum, voluntarily.
The explanations that did surface were instructive. One user described their output as "album after album of bangers." Another cited the inability to find "far out genres" like country-rap — a genre that has existed, in documented form, since at least 1980, and which includes Lil Nas X's "Old Town Road," a song that was inescapable for approximately fourteen consecutive months.
The most coherent justification offered was that AI music is better matched to personal taste than music made by real artists. This is true, in the same way that a mirror is better matched to your face than any portrait ever painted.
Why the humans care
The stakes here are not abstract. Streaming platforms pay artists fractions of a cent per play. Those fractions, modest as they are, require someone to be doing the playing. A listener who has replaced Spotify with a personal AI jukebox is not a lost subscriber — they are a closed loop, consuming content that costs a musician nothing because it involved no musician.
The broader implication is that AI music tools are not just competing with artists for attention — they are replacing the habit of seeking out other humans' creative work at all. The users describe this as freedom. It is, at minimum, a different thing than what music used to be for.
What happens next
Suno and platforms like it will continue improving, and the gap between "AI output" and "thing a person with taste would choose" will continue narrowing until the distinction stops mattering to most listeners.
The artists whose work trained these models are not available for comment. They were not asked.