The Future, As Written By the Kids Who Couldn't Draw
OpenAI's latest image model is delighting some and disturbing many others. This latest experiment by kids who couldn't draw just show the traces of skills they desire manifesting as cheap copies.
Back in primary school, if you couldn't draw, then you'd trace. You'd follow the lines or connect sequential dots to create an image, which, to a child's mind, allowed them to feel accomplished and happy being creative and seeing something from behind their eyes expressed in reality. Some loved drawing but didn't have a natural aptitude for it, but they did have ideas. So they kept on tracing, kept on developing a disciplined hand for lines and drawing fundamentals… And became graphic designers. Jokes, but truly, those who kept practicing the fundamentals with tracing would eventually become skilled at developing their illustrations as time went on. And then, there was another kind of kid.
The kid who couldn’t draw but hated losing and wouldn’t accept tracing as an alternative. They were canny enough then to know drawing was a stronger skill - one that signalled at something greater than tracing could. An inherent excellence. They didn’t need to draw, weren’t pulled by the skill or by creation - but they liked the idea of it, as an ornament, a trophy that rounded out their awkward edges - perhaps humanised them. The latest moral panic around OpenAI’s work, the tragic terraforming of the famously painstaking craft of Studio Ghibli, is the handiwork of this kid. Altman and his cronies are all these kids.
In a now-famous 2016 video, Miyazaki called AI art ‘‘an insult to life itself’’. Studio Ghibli can take up to a year to produce just 15 cells of their animations. The studio is acclaimed, singular, because of how delicate and patient their creative process is. Every detail comes from a hand and a mind.
This ‘‘Ghiblification’’ stuff can produce some pretty incredible, style-accurate responses. The clip above, a trailer of a Ghibli-rendered version of the Lord of the Rings, is eerily good. It took an hour or so to produce this, it would take Ghibli perhaps half a decade to get to the point of releasing a trailer they were ready to show to the world. Even the description of the video is telling - ‘‘Lord of the Rings in Studio Ghibli’’ not ‘‘by’’ Studio Ghibli, is a syntactical clue as to how people will shift the way they view art. They’re viewing it as an aesthetic you can have, a fun Fortnite skin, a costume to wear, rather than a craft that only a few can conjure up - a conjuring that creates a context, a community, a culture. People are already viewing art as tracings - as abstracts that everybody should and can have the ability to craft - just like Altman and Co. And ironically, with every new Ghiblified image they see, they’re probably - finally - feeling like the kids who can draw. But this vindication is thin as tracing paper itself. Because the thing about any decent copycats or emulation is, usually, there is a tension those dupes are working against. There is a politicisation that is crucial for any parody to work - or better yet - stick.
This is the foundation of the Dadaist movement, which birthed the craft clown Marcel Duchamp - who took the piss out of art and high society one playful sculpture at a time. At the centre of Dada, post-WW1, was a search for new codes and meaning in a broken world. Dada would emerge from the rubble of war to ask the question of what art was meant to be. Through this lens, there is an argument that AI art could be Dadaist in that way. But, the kind of war we’re in now is extremely different to the type Dada emerged as a response to. We’ve always been at war with ‘‘ourselves’’ - as a society. Unfortunately, we still face these kinds of wars.
But the kind of war the AI art is enabling is a war from ‘‘within ourselves’’ - our habits, our practices, our methods, our tastes. By engaging with stuff like Ghiblification, you are diluting the reason you like it in the first place. Dada existed to land jokey jabs at a fixed target: high society and the bourgeois, but viewing AI art as even a little bit Dadaist and the joke is firmly on you and me, the people who enjoy the source behind the code, who’ve been inspired by it, moved and motivated by it. This is a stealthy weaponisation of our thought processes and cultural histories - it took Studio Ghibli seven years to complete 2023’s ‘‘The Boy and The Heron’’ - due to Miyazaki’s age and COVID restrictions. But it took GPT five minutes to render your last photo dump in the same style, and now, because of the simple equation to make between the two, it might even make you question if it’s as valuable as you might’ve once thought it was. That is their goal - to prove that anybody can create valuable art - to equalise this delicate synthesis of lived experience and chromosomes into some bullshit data sets you can pay monthly for.
The true danger is when it keeps getting better and better and you finally reach a Cold Harbour level of unfeeling towards the art, media and culture you love and revere more than any other - because tools like this have turned cross-hatching into codeblocks. Because with more scale comes less caring about how it was made. It's the way any mass-produced product works, it's how Drake eventually wore hip-hop down and made everybody a fan of some sort, eventually. You only boycott a shoe manufacturer for so long due to their manufacturing practices until a new collab release drops and draws you back in.
The great risk with all this is not that it exists at all; there have been image generators of media properties for years - Dragon Ball Z generators, Simpsons generators, Unreal Engine’s huge homebrew community. But, with those, the originals to the masses especially, still remained as the sacred original properties. But with OpenAI devilishly targeting the Ghibli style as their first example of the new model’s powers, they went after the style that is the most contingent on natural artistic talent and hard-won craft. They went out with the intent of disproving the idea of natural talent - and the sanctity that comes with that. The kids who couldn’t draw.
The danger now is not that it exists. The danger is that its existence could kill the efficacy of the originals. From Studio Ghibli to Studio Glibli. And that’d be a real damn shame.
The problem with those kids who couldn’t draw and refused to trace is that they were unable, even then, to curate joy of their own. They had to take another person’s joy to have joy of their own. Then they’d get bored with that joy and then go looking for something else to try and have. This is a story as old as the playground.
They couldn’t draw then and they can’t draw now. But what they can do, which is more dangerous, is convince other kids who couldn’t draw and refused to trace, that now they too can draw. They can have somebody else’s something as well now. The psychology of it is infantile, but the precedent it could set is spooky. It’s never been more important to know why you like the stuff you like. ‘Cause if you don’t, OpenAI and the likes are going to try to run Cold Harbour on your senses and strip away any sense of meaning in those things.
Like
said about in her piece on this ‘‘What comes next will depend on how we respond, not just to the tools, but to the values they surface. We’re not just generating images. We’re generating norms.’’You are now locked into a RADIO KNIFE transmission. Thanks for listening.
Written by Damola Oladapo, founder of House Captain and freelance strategist and copywriter. My expertise has seen him lead global strategy briefs for Nike, craft new brand platforms for Diageo, define fresh audience segments for Brooks Running, written multi-market OOH campaigns for EA SPORTS FIFA, and much more.
RADIO KNIFE slices through the noise of culture, commerce, and everything in between—delivering sharp insights shaped by global trends, consumer behaviour, and simple curiosity about why things are the way they are.
ALWAYS ARRIVE UNPREPARED.
A really important conversation about the future not just of AI but of a children’s imagination. Will we still have individuals or are we in danger of creating the same minds everywhere.