Hand it to Oliver McCann for disarming honesty. “I have no musical talent at all. I can’t sing, I can’t play instruments, and I have no musical background at all,” he said, in a recent interview with Associated Press following the announcement of his record deal with Hallwood Media. Why would a music company want to sign on a self-confessed talentless talent? Turns out, McCann has a genius for using artificial intelligence (AI) tools to create music that a lot of people apparently listen to; one of the tracks he created under the stagename “imoliver” has reportedly been streamed 3 million times.
In the three years since ChatGPT’s public debut, the loud laments and dire prognostications about what AI means for art may not have quite receded into the background, but there certainly appears to be a growing ease with the technology. That new artists like McCann would embrace AI to make their dreams of creation come true was only inevitable — one of the most enduring arguments for the technology, from the start, has been its potential for lowering the barriers to entry into the hallowed fields of art.
For McCann and others, AI’s greatest appeal lies in its democratising character: They don’t require an expensive music education; they’re not handicapped by growing up in non-musical environments with limited access to professional/experienced musicians; they don’t need to network and wait outside producers’ offices, hoping for a break. And they absolutely do not need to put in the proverbial 10,000 hours that Malcolm Gladwell posited are required to achieve mastery over any skill.
It also makes sense that music companies are interested in signing on creators who use AI to produce music. Artists who don’t require any songwriting, musical or production support at all are far less expensive than those who do. It could be argued that it makes no sense to do this, as only flesh-and-blood artists, whose music emerges from lived experience and human genius, can rise up the charts and bring in major money, but that doesn’t account for two harsh truths about the music industry. One, at any given time, labels only make money off a handful of their artists. The rest of the artists in a record company’s stable are there either because they sounded promising at one point and the label is waiting for the right moment to offload them, or because they bring artistic cred and reputational heft or because they help diversify the company’s portfolio.
Two, even the biggest of hitmakers can stumble and fall, because who can really predict how tastes change? Take the case of Katy Perry, who spun record after platinum record in the 2010s. Less than a decade on, in 2024, her album 143, remained on the charts for all of two weeks and is reported to have sold just about 500,000 copies (which is also disputed) — a far cry from the 1,000,000 (platinum), 2,000,000 (multi-platinum) and 10,000,000 (diamond) certification levels that Perry once achieved. The point of this digression is that artists capable of constant reinvention are few, and the vast majority of stars are destined to become has-beens.
And what of the listeners? What do we gain or lose as music created by AI sneaks into our streaming playlists? In 2025, we are long past the age of appointment listening. Where once we would make an occasion out of listening to music — ensuring we have access to a player (gramophone, Walkman, stereo system), choosing the right company (or going at it solo), picking albums or songs that match a mood or moment, or catching a regularly-scheduled programme on the radio or on TV — music is now, for better or for worse, woven into the warp and weft of our lives. We listen to it all the time — on the commute, in the gym, at work or in the library, at bedtime. Most of the time, we aren’t even really listening, as the music unobtrusively plays in the background. It fills a silence or becomes the “pink” or “brown” noise, which allows our minds to focus.
This role of music — as undemanding filler — is almost entirely a product of the digital age, where all the user (not listener) needs to do is go to an app, select a mood or song and then allow the streaming service to play track after track, with little or no human intervention. And it’s hardly surprising that much of this music, found in playlists with names like “focus”, “study”, “background”, “reading” and “sleep”, is defined by a pleasant blandness with almost no features that can help distinguish one track from another.
How much of this music is made by humans? In April, French streaming platform Deezer revealed that up to 18 per cent of the content uploaded to its platform everyday is AI-generated. Spotify, which leads in subscriber numbers, has not made its data public, but it has increasingly come under criticism for allowing almost entire playlists to be populated by what appear to be AI-generated tracks. Users have also noted multiple instances of the same track appearing under different “artist” names — a dead giveaway of AI-generated content. The controversy over AI artists on Spotify came to a head in July, when it was discovered that a viral band named The Velvet Sundown, with two albums and over a million streams, was AI-generated, from the music to the promotional images of the band, its backstory and album art.
There is no doubt that the ease of making music using AI can make an already difficult business more cutthroat for artists who don’t embrace the technology. The fact that most listeners still gravitate towards human artists must offer some assurance. Yet, the threat that AI poses is not merely to artists and their livelihoods — it is to the very meaning of art, or what we think of as art. Over the centuries, as the definition of art evolved with new mediums, values and ideas, what remained unchanged, more or less, is the understanding that any work worthy of being described as art demands a lot from the artist. The creation of art is meant to be a journey, one that is marked by the struggle of some kind, whether mental, physical or emotional. It is a labour of making meaning out of experiences and ideas, neither of which AI has, at least at this point in time.
AI can mimic what is palatable about art — the sweetness of a note or the sureness of a rhyme — and it can ensure that the final product is smooth and finished. But art, as we still understand it today, emerges out of the friction of the palatable and unpalatable, the sweet and the bitter. The subjectivity required to make harmony in spite of disharmony — rather than erasing it altogether — still eludes the machine. When that changes, art as we know it may well cease to exist.
pooja.pillai@expressindia.com
