AI won’t kill music. It will drown it.
Most forecasts won’t say it plainly: not a world where artists are replaced, but one where they simply can’t be found. The threat isn’t extinction. It’s noise.
The flood is already here
Spotify already receives roughly 120,000 new tracks every single day. That number has been climbing for years, well before AI entered the picture. Spotify’s own data already shows how hard it is for new artists to break through a catalog that size. Now the cost of producing a release-quality track has dropped to near zero. Tools that once required a studio budget can be prompted into existence in minutes.
Goldman Sachs projects the recorded music market will more than double to over $160 billion by 2030 — driven in part by the sheer expansion of streaming catalogs. More listeners, more markets, and a lot more music. Add an order-of-magnitude jump in AI-generated upload volume and the math for discovery gets worse, not better.
The problem isn’t bad music
Here’s what gets missed in most AI-and-music conversations: the music won’t be bad. That’s the problem. Trained on decades of recorded music, these models produce output that is technically proficient, dynamically balanced, and emotionally inert.
AI doesn’t create culture — it averages it.
IFPI’s Global Music Report consistently shows listeners spending more time on music than ever. But the mechanism for finding something new — algorithmic discovery — is already strained at current upload volumes. A significant increase doesn’t solve that. It collapses it.
The problem isn’t bad music. It’s music with nothing at stake — technically correct, emotionally interchangeable, made for no one in particular.
When everything sounds good, nothing stands out
For most of recorded music’s history, production quality was a moat. You needed skill, equipment, and time to clear a basic bar. That moat is gone. When everyone has access to the same tools, the tools stop being the differentiator.
When everything sounds good, nothing stands out.
What replaces production as the moat? Taste. Identity. Point of view.
You can generate music. You still can’t generate taste.
The thing that makes a listener feel like this song was made for them specifically — not for a demographic, not for a playlist slot. AI has no stake in anything. No loss to write from. No listener it’s accountable to. MIDiA Research has tracked the fragmentation of music consumption for years — that trend accelerates sharply when catalog volume explodes. Understanding how identity translates into real audience growth is exactly what a coherent music marketing strategy has to address now.
Attention was always scarce. It’s about to become the only thing that matters.
What comes next
Five things to expect as this plays out:
Algorithmic discovery gets worse before it gets better. Platforms aren’t built for the volume coming. Recommendation engines will be polluted by machine-generated catalog well before anyone builds tools to filter it. If algorithmic reach is your only growth channel, that’s a fragile foundation.
Trust signals become the new currency. In practice, that means three things: a visible identity that accumulates across platforms (a face, a voice, a consistent perspective); a creative process with enough legibility that your music has a traceable origin; and a small, specific community that shows up. Passive streams won’t cut through. Active proof of life will. What that actually looks like in practice is more concrete than most artists expect.
AI can replicate a genre. It can’t inhabit one.
Niche depth beats broad appeal. The narrower your real connection — the more specific the corner of experience you speak to — the harder you are to replicate. Targeting everyone is now the riskiest strategy. The scene test is a useful lens here: if your music doesn’t place a listener somewhere specific, it gets lost in the catalog.
Curation becomes as valuable as creation. The ability to identify what’s worth hearing — and say why — is already scarce. It will be rarer still when volume is infinite. Artists who can articulate their influences, context, and creative intent have a compounding advantage with curators and listeners alike.
Artists with real identity will widen their lead. Not because the market gets easier. Because the contrast gets sharper. Average is everywhere. Specific is not.
What artists should do now
This isn’t a moment to wait and see. The structural shift is already in motion. Six things that matter more now than they did two years ago:
Stop chasing release volume. One release with real intent and a community behind it outperforms ten releases into silence. Frequency is a tool. Strategy is the point.
Own at least one direct channel. An email list, a regular newsletter, a community you control — something that doesn’t depend on a platform deciding to surface you. Streaming payouts aren’t getting better — that channel is your floor when algorithmic discovery degrades too.
Make your creative process legible. Not a vlog. A visible perspective: why you made this, what you were reaching for, what it cost you. Process content converts passive listeners into invested ones.
Get specific about who you’re actually for. Not a genre, not a demographic — a specific kind of person in a specific moment. Defining your artist identity with that precision is the foundation everything else builds on — the core of what artist development addresses. If your lyrics could have been written by anyone, they will be. Specificity is the last genuinely defensible position.
Use AI tools without becoming one.
Production, arrangement, mixing — use what’s useful. But don’t let tools make your creative decisions. The differentiation isn’t whether you use AI. It’s whether your work has a point of view that couldn’t have come from a prompt.
Build curator relationships directly. Algorithmic playlisting is unstable. Human curators — playlist editors, music supervisors, bloggers, other artists — still operate on trust and taste. A handful of those relationships compound over time in ways no algorithm replicates.
The takeaway
This isn’t a doom scenario for artists who have been doing the work. It’s a leverage point.
The artists who cut through won’t be the ones who sound the best. They’ll be the ones who mean something specific to someone specific — a presence that can’t be averaged, templated, or prompted into existence. Volume is no longer a proxy for value. Distinctiveness is. That starts with understanding why most listeners don’t care — and then giving them a reason to.
If any of this describes where you’re headed, that’s worth a conversation.
Not in the way most people imagine. AI doesn’t write from experience, loss, or identity — it recombines what already exists. The risk isn’t replacement, it’s dilution: a flood of competent, interchangeable music that makes it harder for human work to be found.
Stop competing on production quality — that moat is gone. The advantage now is everything AI can’t fake: a genuine point of view, a consistent identity, a community that has a reason to care. Those things take time to build, which is also why they’re hard to replicate.
Technically, yes — increasingly so. But “good” in the sense of well-produced and sonically competent is not the same as meaningful. AI music optimises for average. It doesn’t have a perspective, a stake, or a story. That gap is the opening for artists who do.
Some are beginning to flag or gate AI-generated content, but enforcement is inconsistent. The more likely outcome is that platforms develop better filtering and discovery layers over time — but that takes years. In the near term, the burden falls on artists to build identity signals strong enough to cut through.
Build what AI can’t replicate: a specific identity, a direct audience channel you own, and a creative process that’s visibly human. Stop competing on production quality — that moat is gone. Own a direct channel (email list, community), make your creative process legible, and get ruthlessly specific about who you’re actually for. The artists who cut through won’t be the ones who sound the best — they’ll be the ones who mean something specific to someone specific.