By Patrick Fasolis
With the release of ChatGPT’s new 4o image generator, users have been using the new tech to turn pictures into Ghibli-style anime frames, despite going against everything the studio and their films stand for. Because of this new technology, the iconic animation style, developed by visionary Hayao Miyazaki, has been diluted into a fun social-media trend. To make matters worse, AI image generators are trained on real artists’ work, mostly without their permission, to create quick, pale imitations of what took some people and teams months, if not years, of hard work.
In The Winds Rises, one of Miyazaki’s films released in 2013, there’s a 4-second shot of a bustling crowd which took the team of animators exactly one year and three months to create. AI models will take hours of footage and other people’s work, and generate a soulless imitation in seconds, but isn’t the beauty of these films the fact that they are labours of love? Or, more broadly, can AI truly “generate” an artistic piece?
The supposed democratisation of art
AI defenders and stakeholders will always rush to say that AI art will “democratise art”, by making it accessible to anyone. This way of looking at art is extremely misleading; democratising anything is always a good thing, which is why saying that AI will “democratise art” seems like a just and noble cause, however art does not stem from privilege nor does it require a specific entry level. From Michelangelo to Pollock, Charlie Chaplin to Stanley Kubrick, there is no one way to make art, it requires a person to freely express themselves, hence it can take on many forms. Art, thus, must be made by a person with a specific intention.
It disingenuous to claim art isn’t already “democratised”, anyone can make art, the hard part is making art that other people will like. Nevertheless, errare humanum est and great art is never made on a whim, it’s the product of days, months, years of hard work, of mistakes and learning from those mistakes. The greatest artworks are labours of love, not convenience, therefore making something that will trivialise such effort is not only anti-art but almost inhuman.
The dangers of AI images
The risk of this new image generator, created by ChatGPT, isn’t merely to create soulless art, often stealing from other people, it could also be used as a tool for fraud or worse. According to an article by TechCrunch, users on X (formerly Twitter) shared how they managed to create near perfect fake receipts.
Such technology can easily be manipulated by bad actors to scam not only businesses but people alike. According to users, this new image generator model is capable of making near perfect pictures of people as well. Gone are the days of 6-fingered AI “people”, it’s now a lot more difficult to discern whether or not an image is real or fake. This technology could easily be co-opted by those who wish to spread hate or misinformation, or worse it could become a tool to create fake revenge porn, ruining people’s lives.
A spokesperson for ChatGPT, when asked why the company decided to make something that could be so easily weaponised by bad actors, claimed that it was to “give users as much creative freedom as possible”. It’s hard to see how the supposed benefits outweigh the risks.
Propaganda’s best friend
Art and propaganda have gone hand in hand since the beginning of mankind. From Virgil’s Aeneid to George Orwell’s Animal Farm, the fact artwork is often an expression of ideas or feelings lends itself well to further political agendas.
However, it takes time to create a work of artistic propaganda and, oftentimes, cracks in the messaging can be found due to the artist’s personal ideals. Virgil did write the Aeneid to praise and legitimize Emperor Augustus, because he had to, but it didn’t stop him from hiding little criticisms in his work. Animal Farm was co-opted by the CIA during the Cold War as a form of anti-soviet propaganda, however the story is critical of fascism as a whole and can easily be applied to many other nations.
There is always nuance in art, even when it’s meant as propaganda, there are always cracks where the truth can shine through. AI image generators are lifeless, soulless even, they’re capable of producing a massive amount of misleading images to shape people’s perception however they like. Recently, the official White House twitter page posted an AI generated image of a woman being deported by ICE in the aforementioned Studio Ghibli style. The image is degrading and trivialises what is a real issue in the United States at the moment: mass deportation.
With reports coming from the U.S. of citizens unfairly questioned by ICE because of the colour of their skin, images like this are extremely harmful. By trivialising what is going on and painting a picture of a white army man arresting brown people as they openly weep, the White House is openly saying that they are unashamed of the brash tactics they’ve adopted.
With fascism on the rise in all Western countries, the ability to churn out hundreds of fake images a day should be cause for concern. The past ten years have been profoundly influenced by the rise of fake news and distrust in Governmental agencies and academic professionals. It’s important to keep fighting for what is true and reject tech that will only benefit billionaires and bad faith actors. To conclude with a quote by the man himself, Hayao Miyazaki, writer and director of some of the most famous Studio Ghibli films, who developed his signature animation style over decades of work, AI art is nothing more than “an insult to life itself”.
Debora Pasero 