If Talent Borrows And Genius Steals, What’s The Problem With Generative AI?
A team from University of Chicago are trying to stop robots from stealing human creativity. Strangely, history is not on their side. Data Universe has been looking at why, and what the solution might be instead.
What is genius? Merriam-Webster says an “extraordinary intellectual power especially as manifested in creative activity”.
I’d like to add “and creative theft”.
Before you think I’m besmirching its good reputation, let’s ask some actual geniuses. Oscar Wilde (The Picture of Dorian Gray) stage-whispered it wittily: “talent borrows, genius steals”. T.S. Eliot (The Waste Land) laid it out lyrically: “immature poets imitate, mature poets steal”. Pablo Picasso (Guernica) said it with both eyes on the one side: “good artists copy, great artists steal”.
And when Steve Jobs (Apple) had designs on the same phrase in California, he was very open about it, saying “we have always been shameless about stealing great ideas”.
Standing on the shoulders of giants
Hip Hop is built on the creative use of samples. Oasis stole from Coca-Cola. The brilliant David Hockney’s latest exhibition – in which he “magically assimilates Picasso’s style” – has been called “a dazzling victory parade”.
All of which is a long way of asking: why is it different for Generative AI?
If genius is “extraordinary intellectual power … manifested in creative activity”. And we know genius steals. Then, ipso facto, Generative AI is a tool that allows us all to touch genius.
So we should be encouraging it to learn and experiment. Right?
AI steals different
Of course Generative AI isn’t Picasso or Hip Hop. But neither is it the brush or the sampler. It can learn, like the artist; it must be guided, like the brush.
It’s a new problem. And it calls for a new solution.
One idea to stop the steal comes from a team at the University of Chicago: they created a tool to “poison” original digital and digitized original artworks so they can’t be used to train generative AI tools. It is, they say: “a last defense for content creators against web scrapers that ignore opt-out/do-not-crawl directives”.
But it feels a lot like trying to stop a rising tide. And we know how that story ends.
What can we do?
I’m not making an argument for mass copyright infringement. Getty Images is suing Stability AI for “misusing more than 12 million Getty photos to train its Stable Diffusion AI image-generation system”. And Getty aren’t the only ones. Everyone deserves credit for their creative work.
But it has been ever thus with light-fingered geniuses. Andy Warhol was sued at least three times by photographers who filed copyright infringement claims against him. And very nearly by Campbell’s Soup – the company only stopping when it realized how much credit Warhol’s works were generating.
Maybe, to steal a little of my own, we need to think different.
Let’s go back, way back
Socrates was a Greek philosopher. Widely credited as the founder of Western philosophy, he died in 399 BC. But not before having his say on the latest technology: writing.
Socrates didn’t believe writing was an effective way to communicate learning. It would, he said, make us all more forgetful when we “cease to exercise memory [and] rely on that which is written”. Luckily his assistant, Plato, wrote down everything he said.
Socrates couldn’t see the future and neither can we. But we have a lot more past to learn from.
Dr. Tomas Chamorro-Premuzic is a psychologist, author and Chief Innovation Officer at ManpowerGroup. He says: “AI is a human invention, so it can only augment our reputation for ingenuity and creativity: creating something more creative than ourselves, makes us more creative as a species.”
We know that trying to stop people from writing worked no better than burning books to stop them from reading. So maybe this time we stop the stopping.
Enhance and reward creativity? That’s the forward-looking solution we’d like to see.
Enjoying our articles? Sign up for our Weekly Insights newsletter to have fresh perspectives and thought leadership from Data Universe and our contributors delivered right to your inbox.
Want Weekly Insights delivered straight to your inbox?
Sign up below to make sure you never miss out on the newest stories from Data Universe.