When OpenAI’s GPT-4o model debuted its multimodal capabilities last month, it sparked a wave of artificial intelligence (AI)-generated art across the Internet. The first viral sensation it triggered was “Ghiblification” — a trending phenomenon of social media users turning ordinary photographs into images evocative of Studio Ghibli’s animation, the Japanese animation studio behind films such as My Neighbor Totoro and Spirited Away.
While global users’ frenzy has put a heavy strain on OpenAI’s GPUs, the trend has raised questions about potential copyright breaches, and pushed discussions concerning artistic integrity and humanity to the forefront in the technology industry.
Many artists have taken issue with AI image generators, arguing that the vast datasets used to train models contain copyrighted works without explicit authorization from creators.
As generative AI tools are known to regurgitate replicas of content they were trained on, any works that these tools create are at best mimicry, regardless of the level of human prompting. This is why so many Ghibli fans found the trend a direct affront to the studio — AI has butchered the cultivated aesthetics of an animation studio famous for its pro-humanistic themes.
Ghibli artists who have spent decades of their lives making transcendent, beautiful animation, are now seeing it “mass produced” or “sloppified” into a meme by a single algorithm in less than a minute.
The traditional time-consuming process of making an animated film, involving hundreds of frame-by-frame images, could soon become a thing of the past. AI replacing human labor is not only a concern for the animation industry, but a hazard for different industries, from art, music, publishing, journalism and law to medicine. With a few instructions and prompts, AI could achieve results humans spent decades perfecting and honing.
It should come as no surprise that Studio Ghibli cofounder Hayao Miyazaki did not find AI an inspiring invention, saying in a 2016 interview: “I feel like we are nearing the end of times. We humans are losing faith in ourselves.”
To strive for better speed, greater efficiency and to cut down on human labor, people have tried to take shortcuts and “outsource” their labor and intelligence to computers. This natural tendency toward sloth could easily turn humans into “cyborgs”: beings that cannot survive or function without machines doing the heavy lifting.
The silver lining so far might be that Miyazaki and Studio Ghibli remain a distinct “style” that acts as an origin or source for AI to be trained. Yet without years of dedication and repeated practice (which AI is now fully capable of achieving after being trained), it is hard to imagine how “style” or the traditional sense of a “master” would be achieved in the future. Without style, humans would lose their unique touch and fingerprint on the world, creating only poor shadows of past artists’ works.
In Taiwan, science, technology, engineering and mathematics majors are widely regarded as being better than liberal arts majors, whose professions are often belittled as either useless or on the brink of replacement by AI. With the proliferation of AI-generated works, it could be expected that fewer people would be working in the arts, and for those who do, AI could soon become a tool that many would rely heavily on for producing works.
The Ghibli AI trend has underscored the conflict between celebrating or tainting a cherished art legacy, and from a wider aspect, between technology and humanity. AI art is a double-edged sword — one that holds incredible promise, but must be handled with care.
It is high time that people find a middle ground where AI art can coexist with traditional art without one undermining the other. To take a step further, it is imperative that humans find a path of not losing faith in ourselves in the world of technology.
