AI Art

The other morning, someone on the Today Programme was getting terribly hot under the collar about AI-generated fashion models. These images, he sniffed, were somehow “fake”, “unrealistic”, even “corrupting”. It was the sort of sermon that could curdle your cornflakes. But the more I thought about it, the more absurd it became. Art has been at this game for centuries.

Take Botticelli’s Venus – a fully grown woman emerging from a seashell the size of a paddling pool. Nobody in 1486 was writing furious letters to The Florentine Gazette saying: “Excuse me, this depiction is deeply misleading – I’ve been to the beach and have yet to see a 6ft blonde drifting ashore on crockery.” Or Monet’s water lilies – lovely things, all dappled light and gentle haze – but they look absolutely nothing like the murky scum ponds we have to fish crisp packets out of every summer.

Impressionism, Cubism, Surrealism – all of them “distorted” reality, and that was precisely the point. Picasso painted faces with two noses and an ear where the mouth should be. AI gives someone legs a foot too long and suddenly the theologians are gasping into their cassocks about civilisation collapsing.

Of course, there is a difference, and it’s one worth admitting. If a person paints a picture or sculpts an object, at least it’s a human interpretation – someone wrestling with the world through their own hands, eyes and peculiarities. Every brushstroke is a trace of thought, every chisel mark a little argument between artist and stone. Even when the proportions are ludicrous or the colours impossible, you can still feel the human intent behind it.

An algorithm doesn’t wrestle. It doesn’t argue. It doesn’t think, it calculates. It doesn’t stare at a blank canvas, despair, drink half a bottle of Chianti and try again. It just… churns. And maybe that’s what really unsettles people – there’s no visible struggle, no hint of the human flaws that make art charming.

But here’s the thing: art has always been an interpretation, not a photograph. Michelangelo’s David isn’t an accurate rendering of the average Florentine lad – he’s a marble superhero who looks like he could bench-press Florence without breaking a sweat. Yet no one shouted “fake!” when he first appeared.

So let’s not pretend this moral panic is about “truth.” It’s just the latest round of “new thing bad.” Once, Impressionism was scandalous; now it’s printed on biscuit tins. Give it twenty years and AI-generated models will be hanging in the National Gallery, while someone on Thought for the Day rails against whatever fresh sorcery has come along to replace them. 

However, if you had the choice of an AI rendition produced by your 5 year old, or a daub done by the same small person, which would you cherish?