AI Art Is Not the Problem—It’s Our Taste
Earlier this month, an AI-generated song climbed all the way to the top of the country charts. Some were amazed by this milestone, which seemed to come sooner than expected. But the dominant reaction was outrage and horror. The thought was that AI is killing creativity. It’s stealing jobs from hard-working musicians. It’s the end of authenticity in music.
The strangest part, though, is what almost nobody asked: Is the song any good?
Because the uncomfortable truth is this: it really isn’t.
Sure, the instrumentation is crisp and the production quality is radio-ready. But the lyrics feel hollow, like a placeholder or template. The hookless melody meanders without surprise. The whole thing sits in a kind of sonic uncanny valley—slick, polished, and utterly lifeless. Think of a musical Polar Express: technically interesting but hauntingly vacant.
Even stranger: there is AI music out there that is compelling, moving, perhaps even great. It just hasn’t hit #1 (yet). So the problem isn’t the tool. The problem is the narrative surrounding the tool. We seem to be caught up in how something is made without bothering to evaluate what it actually is.
When people hear a song was made with AI, the reactions fall almost perfectly into two camps:
- “Wow, that’s awesome!” — even if it’s bad.
- “That’s cheating.” — even if it’s good.
Both takes miss the point entirely. They focus on the process, not the product. And that’s the problem. As a culture, it turns out that judging the product is something we’re becoming remarkably bad at.
The Singer Not the Song
Back in the day, the Rolling Stones released ‘The Singer Not the Song’, and it quickly became an anthem for the bohemian culture of the 1960s. It’s the artist that matters, not the art itself. The principle has a humanistic cachet—it is in the person, not in his or her actions or products, that we find the substance, the meaning, the soul.
But any artist can tell you why this misses the mark. When you’re being transported by a great song or movie or book, what you’re carried away by is not the artists who created it or even the style in which they performed it; you’re carried away by the melody, the cadence, structure of the drama. It might be the case that the singer makes the song cool and share-worthy, but the purpose of the art is sublimity, and you find sublimity in the song.
It’s the same with the culture of today. We’ve built mental scripts—shortcuts, heuristics, badges—that we use to determine whether something is good. Not by examining the thing itself, but by evaluating its pedigree. Someone with an Ivy League degree must be smart. A book agent’s stamp must mean a manuscript is worthy. A celebrity endorsement must indicate quality. It’s come to a point that the inverse is also assumed: You didn’t go to a prestigious college so you must not be smart. You don’t have an agent, so your book must be flawed. No influencers are endorsing your product, so it must be negligible.
Ours is a style over substance culture and the results are unmistakable: Songs, movies, books, architecture, all that seem on the surface to be interesting or compelling but largely devoid of quality. ‘Work’ by Rhianna, ‘Stupid Hoe’ by Nicki Minaj, ‘Sugar’ by Maroon 5, ‘Tennessee Whiskey’ by Chris Stapleton, or anything Taylor Swift has done since 1989.
Take one of the biggest smash hits of the last decade (with 5.6 billion plays on YouTube Music), ‘Uptown Funk.’ It’s pretty obviously not a remarkable song structurally or lyrically. But it nails the James Brown–era style so perfectly that listeners automatically confer greatness on it by association. It reminds them of real excellence—and in our cultural economy, resemblance often counts more than the real thing.
Cultural historian Jacques Barzun warned about this tendency decades ago. In The Culture We Deserve, he describes a society that prizes activity over achievement, information over insight, novelty over excellence. We are busier than ever, more self-conscious than ever, and more obsessed with looking ‘cultured’ than ever—and the result is more often than not a complete lack of cultivation.
Our reaction to AI music is just the newest version of the same script.
People are so fixated on the process—’AI made it!’—that they forget to ask about the product—’Is it any good?’
The scandal with the chart-topping AI country song was not that it wasn’t made by humans, but rather because it was consumed by humans who should have known better. Not because it was threatening legitimate artistry. But because, without even thinking, millions of people streamed it, shared it, and boosted it until it reached #1 as if it were legitimate artistry.
The System Rewards What We Consume
Critics of AI have for the most part arrived at the same conclusion: We must not take part, we must boycott, we must ban all AI generated content. But to ban AI would be like taking the batteries out of the smoke detector to put out the fire. We can’t solve the problem by getting rid of a tool because it’s not in the tool that the problem rests—it’s in the system that prioritizes style over substance and consumers who focus on process over product. If we cheer the singer, not the song, we can’t be surprised when we get celebrity without soul.
This isn’t to say the entire musical ecosystem is broken. There is an astonishing amount of good music out there, and legacy institutions do get it right sometimes. But, as we’ve discovered, the system doesn’t exist to reward excellence. It exists to reward attention. And, increasingly, attention lacks discretion.
That’s why mediocre songs—human or machine-made—go viral.
That’s why derivative sounds dominate playlists.
That’s why a bland AI track can soar to the top of the charts.
Consumers, by and large, are not trained to distinguish between art that merely imitates the shape of emotion and art that actually evokes emotion. Between songs that are competent and songs that are transcendent.
The arrival of AI hasn’t changed that dynamic—it has exposed it.
If we can’t tell the difference between a real song and a hollow imitation, the problem is not the code. It’s our ears. The challenge of this moment isn’t that machines are learning to make art. It’s that we must relearn how to judge it.
