The ChatGPT Studio Ghibli trend: hurtling towards the point of no return
My uncle would get DVDs of his favourite films from the US every time he’d visit. When I was ten, and he came to our place, he showed us some of his animated film picks. The first of these was Studio Ghibli’s My Neighbour Totoro. Before watching this film, I was usually ignorant of cinema, preferring Cartoon Network and Nickelodeon. I’d always thought of movies as boring, and if you’d asked nine-year-old me what my favourite film was, I’d probably have said the 2005 version of Charlie and the Chocolate Factory.
After watching My Neighbour Totoro, something changed. Maybe it was the addictive soundtrack, or the theme song, or the stunning animation. The first thing that got through to me, was however, the message. The universality of it. As two little girls move into a house in the Japanese countryside, their mom is hospitalised. While they seem happy, there are undercurrents of grief. The girls concoct adventure, spend time with this giant but adorable troll, and his little children? Siblings? Deep down—they’re terrified. They just want their mom to come back home.
They’re beautiful, all have stunning soundtracks, effortless magical realism and fantasy, but when you strip everything down, there’s almost always a profound message
I suppose that’s why Hayao Miyazaki and Studio Ghibli movies strike such a chord in me—and in numerous people across the world. They’re beautiful, all have stunning soundtracks, effortless magical realism and fantasy, but when you strip everything down, there’s almost always a profound message. A raw wound, a need for vindication, a childish dream, a blood-curdling fear. Something, anything.
In late March, we were met with a new AI fad. I’d gone back home to India for Easter break, and it was especially viral there. People had taken no time to hop on the bandwagon—many of them not remotely familiar with Studio Ghibli. In seconds, people would give prompts and get an AI-generated Studio Ghibli “inspired” image out of it.
I’m no stranger to AI art—since 2021 with the launch of Dall-E, I’ve seen too much of it. From three-handed, six-fingered people to gratuitously graphic videos of anthropomorphic cats. I’m largely ambivalent towards it, skipping past the videos and trying to not think too much about them. With Studio Ghibli, however, something broke in me. Not just because it was rendering something so original into a cookie-cutter, not the extra fingers it’d add, not even the way it removed people’s freckles, or lightened their skin tones. Something deeper.
Since AI models are trained on pre-existing art, unbeknownst to artists, we feel compelled to ask if AI art is merely a regurgitation of other people’s works
Producer Hirokatsu Kihara revealed in a now-deleted video the process that went into animating a Studio Ghibli film. Having worked in Castle in the Sky, My Neighbour Totoro, and Kiki’s Delivery Service during 1985 to 1989, he revealed how every frame would be hand-drawn, and be critiqued by filmmaker Hayao Miyazaki before being approved. He called the process “unimaginable”.
In 2020, producer Toshio Suzuki shared it took Studio Ghibli “one month to complete just one minute’s worth of animation,” owing to Miyazaki’s aversion towards CGI animation. In a viral clip from 2016, when shown artificial intelligence animation, Miyazaki called it “creepy” and “an insult to life itself”.
Studio Ghibli’s art style is unique in its fluidity. Films such as Spirited Away and the recent The Boy and the Heron feature worlds with large scopes, where several distinct characters must be animated. Since the animation style is consistent across films and at the level of the scene, the animations don’t tend to compromise on quality. Additionally, magical realism is an integral element in Studio Ghibli films, this would also require intricate character designs and backdrops.
Studio Ghibli hasn’t reacted to the situation yet, but the argument against AI art has been widely discussed. Artists and people against AI art take issue largely with plagiarism. Since AI models are trained on pre-existing art, unbeknownst to artists, we feel compelled to ask if AI art is merely a regurgitation of other people’s works.
What is real about a recreation of a man and his dog that gets generated within seconds? What is a work of art without the labour, the thought, and the consideration that goes into making it?
A predictable rebuttal would be that most art is inspired by some art. While I can’t refute this claim, I can say that if someone makes a painting in the style of another artist, and claims they pioneered the style themselves, they’d be met with criticism. When AI outputs that picture of an anthropomorphic cat wearing a top hat, it doesn’t say what inspired it.
With Studio Ghibli it hurts even more because we know what goes into making these films. Countless days, a month for a minute’s worth of animation, immeasurable hearts and souls. A crux. Something real. What is real about a recreation of a man and his dog that gets generated within seconds? What is a work of art without the labour, the thought, and the consideration that goes into making it?
We say we’re scared of AI taking over the world, of it doing what makes humans unique while we work on mundane tasks. There’s a sense of cognitive dissonance that concerns me, then. While I’m not surprised by rich CEOs hopping on the trend, I am shocked at seeing people who may have opposed just this a few weeks ago. I am concerned about the “exceptions” we make, because a trend isn’t that “deep”. I think it is just that deep.
But I missed the memo—I missed the point when we decided we’re okay with letting it get into the humane—into the arts and the humanities.
Back when everyone I knew feared AI in 2021, I calmed myself down. I maintained that people would use it to help themselves, further their ideas, generate algorithms and make strides. I saw it as a tool. A tool to come to conclusions we hadn’t arrived to yet, to hasten compelling innovations. But I missed the memo—I missed the point when we decided we’re okay with letting it get into the humane—into the arts and the humanities.
It is never helpful, however, to just look around and feel hopeless. I think it’s important we think a bit more before we jump into another AI art fad—to wonder if it’s worth it. To wonder what it adds to our lives in the long term. If you truly are against AI in the arts, there ideally shouldn’t be any exclusions in this equation.
I first saw My Neighbour Totoro in 2016—I felt scared when the girls felt scared, I felt fascinated when the girls started going on their woodland adventures, I laughed when their dad was clueless about everything around him. I’ve seen it countless times since. Last year, I saw it screened at the student cinema, and it struck me even more than the first time I saw it. When I was a child, I was scared for their mother, but as a young-adult, I felt the fear of grief. The fear of losing your childhood, of being forced to grow up too soon.
When I saw the first Studio Ghibli AI rendition, I felt a sense of resignation. I knew that a week later everyone would be sick of it. Then, I felt afraid. Afraid that we might be hurtling towards the point of no return—and maybe it’s time to hit the brakes and look around for a second.
Comments