Millions of AI-generated songs are now created daily with no human performer involved. Photo / Getty Images
We’re told that AI was inescapable in pop music this year, but I escaped just fine.
In July, when journalists revealed that the Velvet Sundown, a rising new rock quartet on Spotify, was, in fact, a fake band created with artificial intelligence, I chose not to listen to the songs.
Months later, when an AI-generated act called Breaking Rust topped Billboard’s country digital song sales chart, I kept my fingers in my ears. Same whenever I encountered headlines about the emerging AI gospel entity Solomon Ray, or TaTa Taktumi, the new AI protege of super-producer Timbaland.
Considering the embarrassing amount of money and hype wasted on AI-generated music this year, resistance was easy.
Uncomplicated, too. Like so many who have grown sceptical of AI, I value my life. I don’t want AI-generated music taking a moment of it away from me. Because it isn’t music.
The utopian sales pitch that tech CEOs continue to recite goes something like: AI is an altruistic pursuit that will relieve humanity from the burdens of effort and decision-making.
But music – making it, listening to it – requires effort and decisions, curiosity and contemplation.
Music is something that emerges through experience and develops through tradition, binding us socially in the process. When you listen to your favourite song, do you want to feel like you belong to a community or a dataset?
Creators of the generative AI music platform, Suno, say their product was trained on a dataset of pop recordings numbering in the tens of millions, which allows the platform to work like so: Type a short text prompt describing the kind of song you’d like to hear, and Suno plops one out.
According to Billboard, Suno users are doing this more than 7 million times every day, creating a profusion of brainless songs that seemingly manifest the bleakness of Suno CEO Mikey Shulman’s vision of music itself.
“It’s not really enjoyable to make music now,” Shulman said in an interview in January. “It takes a lot of time, it takes a lot of practice, you need to get really good at an instrument or really good at a piece of production software … I think the majority of people don’t enjoy the majority of the time they spend making music.”
Other apologists say generative AI is just the latest technological scapegoat in pop.
Being against Suno in 2025 is like being against synthesisers in 1977, or against drum machines in 1984, or against Auto-Tune in 2005. But those technologies result in music only when deployed by human decisions – and, as listeners, much of our pleasure still lies in identifying those human decisions.
Since the dawn of Kraftwerk, we’ve enjoyed dancing on whatever line separates the real (acoustic sounds that shake the air) and the fake (digital sounds inside a box). But if we’re listening to music made entirely with generative AI, there’s no guessing game to play, no dancing to be done.
And if sentient pop musicians ever figure out how to collaborate with generative AI in innovative ways, the innovations will be human.
Suno is trained on the history of popular music, which means the history of popular music is all that it can spit back at us.
In his recent book, Against Platforms: Surviving Digital Utopia, technologist Mike Pepi underscores this shortsightedness, arguing: “If AI maximalists have their way, we will eventually run out of training data, confining creative AI to a vicious cycle of ever less impressive output. From this standpoint, AI creativity is not a new frontier but a dead end.”
Living, breathing pop stars are already scoping trouble on those horizons. In November, the Spanish pop futurist, Rosalía, released Lux, an outwardly ambitious album she made with the London Symphony Orchestra, with songs about sainthood and martyrdom sung in more than a dozen languages.
Lux was an elaborate concept album that felt finely wrought and tightly controlled but its grandest statement seemed to be in Rosalía’s proud refusal to use AI in any aspect of its creation. Music this extravagant requires countless decisions, and Rosalía wanted to make them.
Listeners make decisions, too. We can abstain from listening to AI-generated music just as easily as we can cancel our streaming service subscriptions, or boycott Ticketmaster, or decide to start listening to CDs again, or take up the clarinet.
It’s easy to forget the power we hold over our own listening lives, especially when our society presents a technological change as an inevitability.
The CEOs who stand to profit so obscenely off these seismic shifts will scold us for being ignorant and stubborn, warning us that we’re about to be left behind. But they’re headed toward a world without music, and I don’t want to go.
Sign up to Herald Premium Editor’s Picks, delivered straight to your inbox every Friday. Editor-in-Chief Murray Kirkness picks the week’s best features, interviews and investigations. Sign up for Herald Premium here.