In a bitter irony, people around the world are using OpenAI’s GPT‑4o model to generate images in the style of legendary Japanese animator and filmmaker Hayao Miyazaki, who for his entire career has prized values of humanism, painstaking artistry, and the sanctity of nature versus the horrors of exploitative industry. And while neither he nor Studio Ghibli, the Tokyo animation studio he founded, has publicly commented on the meme trend (the company did not immediately return a request for comment from Rolling Stone, either) fans are in an uproar.
“It’s been 24 hours since OpenAI unexpectedly shook the AI image world with 4o image generation,” wrote a tech newsletter columnist in an X post on Wednesday, referring to the Silicon Valley company’s new paid text-to-image model, available only through the paid version of ChatGPT, which costs $20 monthly. Among the examples of what 4o could do, he shared what he called “Studio Ghibli style memes” — that is to say, recognizable memes such as “Distracted Boyfriend” and “Disaster Girl” rendered as anime-like illustrations somewhat reminiscent of Miyazaki’s for beloved classics such as Spirited Away and Howl’s Moving Castle.
While some were impressed, many of the responses were brutal. “I saw the beautiful 4K IMAX version of Princess Mononoke last night… then got out and saw this shit,” wrote one critic, naming a 1997 Miyazaki film that is currently enjoying a limited theatrical run (and certainly has points to make about the environmental destruction wrought by technology). “Do you morons truly value art so little that it’s just a filter for your profile pic? Pay a fucking artist and make something real you gremlins.” Another complained, “People act like this is a good thing but all the soul has been sucked out of society, we really are at the peak, nothing really matters anymore.” A third detractor simply replied: “Hope Studio Ghibli sues everyone involved.”
Indeed, the copyright question became a pressing concern as soon as ChatGPT subscribers began churning out the Ghibli-inspired content and filtering their own photos through the aesthetic. Although OpenAI CEO Sam Altman himself encouraged the craze by changing his X profile pic to a Ghiblified version of himself — and joked about receiving hundreds of messages from people who had sent him similar images — the company had by Wednesday evening implemented a guardrail that made it hard to generate such content. “We added a refusal which triggers when a user attempts to generate an image in the style of a living artist,” OpenAI said in a statement, though a spokesperson told Business Insider that 4o would still be able to imitate “broader studio styles.” Of course, it wasn’t difficult for ChatGPT enthusiasts to figure out prompts that produced Ghibli memes without using the name of the studio. Another problem is that users were able to generate Ghibli-ish images depicting 9/11, the assassination of President John F. Kennedy, and domestic terrorist Ted Kaczynski, the “Unabomber.”
Miyazaki, 84, is certainly a living artist, and it’s not hard to guess what he’d make of 4o. Known for his intensive, perfectionist labor over hand-drawn frames, he is a figure of existing memes already. In a 2016 documentary about his career, he visits with a team that shows him a demo of their AI-generated computer animation of a disfigured zombie character. Miyazaki calmly expressed his withering contempt: “Whoever creates this stuff has no idea what pain is whatsoever,” he said. “I am utterly disgusted. If you really want to make creepy stuff, you can go ahead and do it. I would never wish to incorporate this technology into my work at all. I strongly feel that this is an insult to life itself.” He also observed, “We humans are losing faith in ourselves.” Screenshots of the devastating exchange and Miyazaki’s final verdict have circulated on social media ever since.
One telling indicator of exactly who is interested in pumping out AI slop with ChatGPT: On Thursday, the White House‘s government X account mocked the deportation of an alleged fentanyl dealer by Immigration and Customs Enforcement earlier in the month by rendering her as a crying Ghibli character being arrested by an agent. Meanwhile, Nayib Bukele, the president of El Salvador and self-proclaimed “world’s coolest dictator” — who in President Donald Trump‘s second term has accepted migrants deported from the U.S. to be held in the country’s notorious “mega-prison,” CECOT — shared a Ghibli version of himself on X. Writer and critic Gareth Watkins, among others, has convincingly argued that AI provides “the new aesthetics of fascism,” with right-wing extremists eager to leverage it for lightning-fast propaganda.
But the backlash this time around, even to the apolitical content, has been intense. “Look, call me the fun police or whatever but Hayao Miyazaki didn’t spend his entire life cultivating a distinct artistic identity so that AI could reduce decades of his discipline and integrity to meme fodder,” wrote one X user. “It’s a shame how easily y’all abandon principle just to feel included.” Someone else was more pointed on the subject, responding to a person who had generated a famous 2018 photo of Trump, former German Chancellor Angela Merkel, and other G7 leaders at the time in the Ghibli style: “hayao miyazaki should start killing people. I think.” The post has more than 270,000 likes.
On Thursday, Altman acknowledged that the 4o craze was putting a strain on OpenAI’s resources, noting on X: “it’s super fun seeing people love images in chatgpt. but our GPUs are melting. we are going to temporarily introduce some rate limits while we work on making it more efficient. hopefully won’t be long!” He also announced that the free version of ChatGPT would soon allow for up to three image generations a day. The day before, he appeared to defend the use of AI tech for this purpose, posting, “one man’s slop is another man’s treasure.”
“Says the CEO of Big Slop,” replied a film artist and illustrator.
