The Newest Brush in an Old Studio

A worried writer on AI, art, and why he’s building a business on it anyway

Pull up a chair, and let’s begin with a confession. The first time I heard a song composed by a machine, I felt a small, cold prickle at the back of my neck. It was lovely enough, a wistful piano line that wandered exactly where a melody ought to wander. And that was precisely what unsettled me. If a program can find the ache in a minor chord, what exactly is left for the rest of us?

I’d like to turn that question over with you, because I suspect you’ve asked some version of it yourself. Maybe it came while scrolling past an image no human hand had touched, or while reading a paragraph that sounded almost, but not quite, like a person. Does art made by artificial intelligence cheapen the art made by us?

To answer honestly, I think we have to take a long walk first.

Come with me through an imaginary museum, one that holds every tool humans have ever used to make something beautiful. In the first room the walls are dim and rough, and horses gallop across the stone in charcoal and red ochre. Some seventeen thousand years ago, someone in what is now southern France ground minerals, mixed them with animal fat, and blew pigment through a hollow bone. That was technology, too—the cutting edge of its day.

The rooms beyond move faster. Here is oil paint, which let the Flemish masters render velvet and candlelight with uncanny depth. Here is Gutenberg’s press, which in the 1450s began stamping out pages while scribes worried that books would become common, cheap, soulless. Here is the daguerreotype, at which the painter Paul Delaroche reportedly glanced before declaring painting dead. Painting simply dusted itself off, picked up the new tin paint tube, carried its easel outdoors, and invented Impressionism, freed from the chore of merely recording what the eye could see.

Keep walking. The phonograph was going to empty the concert hall; instead it carried the blues into farmhouses that had never heard it. The talkies were going to ruin cinema; they gave us Casablanca. Writers regarded the typewriter as a vulgar go-between for the mind and the page. Photographers who had mastered the darkroom eyed Photoshop with outright alarm. Musicians grumbled that the synthesizer was cheating, then the drum machine, then Auto-Tune. Graphic designers who set type by hand saw an entire trade fold itself neatly into a laptop, and then a generation of designers grew up whose native tongue was the pixel.

And on a July evening in 1965, Bob Dylan walked onto the stage at the Newport Folk Festival with an electric guitar, and part of the crowd reacted as though he’d brought a chainsaw to a christening.

Notice something as we stroll: in nearly every room, someone stands in the corner insisting this new contraption will be the end of real art. Even Socrates, as Plato tells it, fretted that writing itself would weaken our memories. Each time, something genuinely was lost—a skill, a livelihood, a certain texture. And each time, something new grew in the space left behind. The tools never made the art. They changed who could make it, and how.

Now we reach the newest wing, still smelling of fresh paint. It’s full of images no hand touched, stories drafted by algorithms, songs composed in seconds. And here I owe you a second confession, a more personal one.

I have spent most of my writing life making things up. Fiction is where I live: characters who refuse to behave, sentences rewritten nine times at two in the morning, the slow, stubborn work of discovering what a story is actually about. So when I watch a machine produce a passable short story in the time it takes me to refill my coffee, I don’t feel like a detached observer of history. I feel it in my chest. It worries me. Not because the machine’s story is better—it usually isn’t—but because “passable” and “instant” make a powerful pair in a world that doesn’t always have patience for “good” and “slow.” I worry about the young writer who never gets to finish her clumsy first novel because the market for clumsy first novels has been flooded. That worry is real, and I won’t wave it away.

Which makes what I’m about to tell you a little awkward. I’m starting a company built on AI: an AI-driven marketing firm that helps small businesses use these very tools to tell their stories, reach their customers, and do in an afternoon what once took a week.

You might reasonably ask how a worried author ends up there. The honest answer is that worry and wonder have been sitting side by side in me for a while, and I decided the wiser course was to understand this thing from the inside rather than shout at it from the corner of the room. The dry cleaner down the street, the family bakery, the hardware store that has outlasted three chain competitors: these businesses have never been able to afford an agency, a copywriter, and a designer. AI puts something like that within reach. That’s not nothing. That’s a door opening for people who have been standing outside it.

Working with these tools every day has also taught me where their limits are, and those limits are exactly where my hope lives. A machine can generate a thousand slogans, but it cannot tell you which one is true to the woman who opened that bakery with her grandmother’s recipes. It can produce a thousand paintings, but it cannot tell us which one matters, or why. That judgment remains ours. AI does differ from earlier tools in one important respect: the old ones waited for instruction, and this one can volunteer. Still, volunteering is not the same as meaning something.

So does AI-generated art cheapen art? It certainly cheapens some of it in the plainest economic sense. An illustration that once took a week now appears in seconds, and that is a real loss for real people. But the worth of art has never rested solely on the difficulty of its making. We cherish a poem because someone meant it—because a particular life pressed against the language until it gave. When anything can be made effortlessly, effort stops being the measure of value, and intention takes its place. We’ll ask less “How hard was this to make?” and more “What was this person trying to say, and did they say it?” A machine can supply the brushstrokes. It can’t yet supply the reason.

Here, I suspect, is where AI may find its place: not as the artist, but as the newest brush in an old studio. The painter still decides where the stroke goes. At the writing desk and at the new business alike, that is the line I intend to hold. The tool drafts; the human decides. The tool suggests; the human means it.

Will that line hold for all of us? Honestly, I don’t know, and anyone claiming certainty is selling something. Perhaps this new wing will one day look as natural as the others, one more room on the long walk. Perhaps it will change the museum entirely. Time will tell.

What time won’t do is run backward. The genie is out of the bottle, the crowd can boo all it likes, and the music, as it always does, will go on. Our task now is to walk into that newest room, look closely, and teach the genie some manners.