Jun 2025 · 9 min read

Does 'new' music matter?

Streaming taught a generation to listen without caring when a song was made. Generative AI is now teaching them to listen without caring who — or what — made it. That shift, from era-agnostic to context-agnostic listening, is quietly rewriting the economics of being a working artist.

The timeline collapsed

For most of recorded history, popular music moved in one direction. To listen was to keep up. Falling behind meant being culturally marooned, left defending a single the rest of the world had moved past.

That logic no longer holds. Platforms like TikTok have flattened the catalogue, surfacing a song from 1983 alongside a release from last Thursday with no hierarchy between them. The cultural result is a generation of genuinely era-agnostic listeners — playlists that move without friction from The Beatles to Nirvana to Chappell Roan, connected by nothing but the algorithm that surfaced them.

For a while, that freedom looked like an unambiguous good. The complication arrived from a different direction.

The upload flood

On paper, the streaming economy is healthier than ever. Spotify paid out a record $11 billion in royalties in 2025. [3] But the pool is finite, and it's now being drawn down by content no human made.

In late 2025, Deezer reported that fully AI-generated tracks accounted for roughly 39% of its daily uploads — about 60,000 songs every 24 hours, up from around 10,000 at the start of the year. [4] CISAC projects that generative AI could put 24% of human creators' revenue at risk by 2028 as synthetic output competes for the same finite royalties. [5] In a single year, Spotify removed 75 million tracks it classed as "spammy." [3]

When the 'who' stops mattering

Era-agnostic listening was, at worst, indifferent to when a song was made. The newer mode is indifferent to who made it. Music is increasingly received passively — saved without a glance at the artist, discovery handed to the feed. Roughly three-quarters of Gen Z now find music primarily through platform recommendations. [7] The habit of connecting a sound to a person has quietly eroded.

That indifference is what makes the present moment precarious. If a listener doesn't care whether a track came from a human or a machine — and the machine can upload at a scale no human could match — the basic arithmetic of an artist's career stops working.

The stakes are not abstract

Recorded music and live music are a single economy. Streaming rarely sustains an artist on its own; live performance does the heavy lifting, and live performance depends on an audience that cares enough to show up. Festivals fold when ticket sales don't cover the lineup. Local venues close when no one books the room. Both rely on listeners paying enough attention to choose a human artist over an infinite, frictionless alternative.

So the question — does new music matter? — turns out to be economic as much as aesthetic. Attention is a form of spending. Where audiences direct it decides who gets to keep making.

None of this is an argument for nostalgia. It's an argument for deliberate curiosity: attention directed, on purpose, toward the artists, labels, and curators still thinking carefully about what gets made and why. Listening is, among other things, an economic act — and in a market flooded with synthetic supply, choosing where it goes may be the most consequential thing a listener can still do.