From Gimmick to Flood
What started as an experimental curiosity has quietly become one of the most disruptive forces in the music industry. Generative AI — once the domain of academic researchers and avant-garde artists — is now producing tracks at a scale that streaming platforms were never designed to handle.
The early signs were intriguing rather than alarming. In 2018, artist Taryn Southern released I AM AI, an album created with significant AI assistance. A year later, Holly Herndon's Proto pushed the idea further, exploring what it might mean for a human artist to train and collaborate with an AI model. These were genuine artistic experiments, and they were received as such: curious, sometimes unsettling, but clearly the work of people wrestling with new tools.
The Industrialization of AI Sound
That spirit of experimentation didn't last long before the economics took over. Tools like Google's Magenta made AI music generation more accessible, and a wave of startups followed, each promising to lower the barrier to producing passable — and then surprisingly decent — background music, stock tracks, and eventually full songs.
The result is a quiet flood. Streaming services are now home to millions of algorithmically generated tracks, many uploaded in bulk by accounts with no clear human artist behind them. Some of it is ambient filler — the kind of lo-fi study music or sleep sounds that rack up passive streams. But increasingly, AI is being used to generate content that mimics real artists' styles closely enough to compete for listener attention.
What Listeners Actually Want
Here's the uncomfortable question the industry keeps circling: does anyone actually want this?
The data is mixed. Passive listening — playlists, background music, algorithmic radio — seems indifferent to origin. If a track sounds right for the mood, listeners may not care how it was made. But active listeners, fans who follow artists and seek out new music with intent, appear far more skeptical. Authenticity remains a core part of how people form attachments to music, and the story behind a song still matters.
Spotify, Apple Music, and other major platforms have been slow to develop clear policies around AI-generated content. Labelling is inconsistent. Discovery algorithms don't distinguish. The effect is that human artists are competing for playlist placement against tracks that cost essentially nothing to produce.
The Industry Reckons
Record labels, songwriters, and performing rights organizations are pushing for clearer rules — on disclosure, on royalties, on what counts as artistic authorship. Legislation is lagging. In the meantime, the volume of AI-generated music keeps growing.
For musicians, the worry isn't that AI will replace the best artists — it's that it will devalue the middle tier, the working musicians who built sustainable careers on sync licensing, streaming royalties, and consistent output. Those are exactly the kinds of tracks AI can now replicate cheaply.
The question of who benefits from AI music — and who gets hurt — is still being written. But the flood is already here.
Source: The Stepback newsletter via The Verge
