Everyone knows the song “Video Killed the Radio Star” by the Buggles, it was one of the earliest big hits on MTV (back when it was still called Music Television). The song is pretty good, but it also speaks to a genuine fear and wonder about our world, that changing technology upends our social fabric and destroys our livelihood. The radio star who just wasn’t pretty enough for video, or couldn’t compete with the big production values of music videos, or just didn’t like dancing and being seen at all. That radio star is the Dickensian protagonist of the modern age, as they are tossed aside and replaced when new technology comes along.

This Luddite fear has pervaded throughout history. The loom-smashing followers of Ned Ludd are only the most famous, but there were silent actors who never made it in talkies. There were photo-realistic painters who could never compete with a camera. John Henry died trying to beat a steam drill. In each case, an argument could be made that the new technology removed some important human element. The painters could claim that photography wasn’t “true art”. And the loom smashers too probably believed that their handcrafts were more “real” and more deserving of respect than the soulless cloth that replaced it.

So why is AI art any different? Why should we care about the modern Luddites who want to ban it or restrict it? I say we shouldn’t.

AI art steals from other artists to make its images

common argument

No more than any artist “steals” when they learn from the old masters. It is a grievous misunderstanding of how AI works to claim that it cuts and pastes from other images, and an AI training itself on a dataset of art is no different than an art student doing the same whether in university or on their own. The counter-argument I’ve heard is “why are you ascribing rights to an AI that should only belong to humans! Yes humans can learn from other art, but AI shouldn’t have the right to!” I’m not ascribing anything to AI, the person who coded the AI and the person who used the AI have the right to use any images they can find, just as an artist does. And just as the output of an artist learning from old masters is itself new art, so too is the output of coding or using an AI that has been trained on old works.

AI art is soulless

common argument

As soulless as loom-made fabric is compared to hand-made. Or as soulless as a photograph is compared to a hand-painted picture. Being made with a machine doesn’t detract from something for me, and I think only bias causes it to detract from others.

AI art takes money out of artists’ pockets, it should be banned to protect the workers’ paychecks

common argument

Why is the money of the workers more important than the money of the consumers? Loom-made fabric competes with hand-spun fabric, should we smash looms to keep the tailors’ wages up? Are we ok with having everything cost more because it would hurt someone’s business if they had to compete against a machine? The counter-argument I’ve seen to this is that the old jobs replaced by AI were all terrible drudgery and it’s good that they were replaced, whereas art is the highest of human expressions and should never be replaced. Again I think this is presentism and a misreading of history. I’m sure there were tailors and seamstresses who though sewing and making fabric was the absolute bomb, who loved their job and though that their clothes had so much heart and soul that they were works of art in and of themselves. And I know there are artists in the modern day for whom most of their work is dull drudgery.

Thinking that your job and only your job is the highest form of human expression and should never be replaced, well to me that just shows a clear lack of empathy towards everyone else on earth. No one’s job is safe from automation, but all of society reaps the benefits of automation. We can all now afford far more food, more clothing, more everything, since we started automating manual labor. Labor saving creates jobs, it doesn’t destroy them, it frees people to put their efforts towards other tasks. We need to make sure that the people who lose their jobs due to automation are still cared for by society, but we should not halt technological progress just to protect them. AI art allows creators and consumers to have more art available than they otherwise would. Game designers can whip up art far more quickly, role-players can get a character portrait without having to pay, this lets people have far more art available than they otherwise would. In the same way that the loom let us have far more clothing available than we otherwise would.

AI art is always terrible

common argument

I find it funny that this often comes paired in internet discourse with “I’m constantly paranoid and wondering if the picture I’m looking at was made by AI or not.” There’s a very Umberto Eco-esque argument going on in anti-AI spaces. AI is both terrible and easily spotted, but also insidious and you never quite know if what you’re seeing is AI, and also everyone is now using AI art instead of “real” art.

If real art is better than AI art, wouldn’t there be a market for it still? There’s still a market for good food even though McDonald’s exists, if AI art is terrible and soulless than it isn’t really a danger to anyone who can’t make good art themselves. And if AI art is always terrible, then why are so many people worried about whether the picture they’re seeing is AI-made or not? Shouldn’t it always be obvious?

This is very obviously an emotional argument. If you can convince someone that a picture was not made with AI, they’ll defend it. If you convince them it was made with AI, they’ll attack it.

This was a vague disconnected rant, but I’ve become sort of jaded to the AI arguments I’ve seen going on. I had thought that modern society had somewhat grown out of Ludditism. And to be frank, many of the people I see making anti-AI arguments are supposedly pro-science and pro-rationalism. But it seems that ideology only works so long as their “tribe” doesn’t ever get threatened.