OpenAI has been forced to introduce temporary rate limits on ChatGPT’s image generation feature due to overwhelming popularity, according to CEO Sam Altman. The surge in usage has put significant strain on the company’s GPU resources.
“It’s super fun seeing people love images in ChatGPT. But our GPUs are melting,” Altman shared on X. “We are going to temporarily introduce some rate limits while we work on making it more efficient. Hopefully won’t be long! ChatGPT free tier will get 3 generations per day soon.”
The company launched this native image-generating capability, powered by the GPT-4o model, on Tuesday. Altman described it as a “new high-water mark for us in allowing creative freedom,” noting that the tool produces more realistic images and improves text rendering.
While Altman didn’t specify the exact new rate limits, the situation highlights the immense computing resources required for AI image generation. Generative AI is extremely compute-intensive, requiring powerful GPUs, large-scale cloud infrastructure, and significant energy consumption to process the vast number of calculations needed.
Users have primarily been using the new tool to create AI-generated content in the style of Studio Ghibli, the beloved Japanese animation studio known for films like “My Neighbor Totoro” and “Spirited Away.” Social media platforms have been flooded with Ghibli-style representations of people, animals, historical events, and even podcasts.
Altman himself joined the trend, updating his X profile picture to a Ghibli-style version of himself.
The flood of Ghibli-style images has reignited debates about potential copyright violations. Many artists have criticized AI image generators, arguing that the vast datasets used to train these models contain copyrighted works without creators’ explicit permission. OpenAI is currently facing several copyright lawsuits to this issue.
In what appears to be an effort to address some of these concerns, OpenAI stated in its system card for the 4o image generation that it had “added a refusal which triggers when a user attempts to generate an image in the style of a living artist.”
However, this policy raises questions in the case of Studio Ghibli, as its co-founder Hayao Miyazaki is still alive. Miyazaki has previously expressed strong views against AI, famously calling it an “insult to life itself” in a 2016 documentary.
The free version of ChatGPT reportedly refuses to generate images in the style of Studio Ghibli, citing policy guideline violations. When Fortune magazine prompted the chatbot to create an image in the Ghibli style, it responded: “I wasn’t able to generate the image because the request didn’t follow content policy guidelines. If you’d like, I can create something similar with a different approach—perhaps a symbolic representation of technology and art clashing in a fantasy setting. Let me know how you’d like to proceed!”
When asked about this apparent contradiction, OpenAI told TechCrunch that while ChatGPT refuses to replicate “the style of individual living artists,” the company does permit it to replicate “broader studio styles.”
OpenAI representatives did not immediately respond to requests for further comment on this issue.
The surge in popularity demonstrates both the impressive capabilities of these new AI tools and the complex ethical questions they continue to raise as they become more accessible to the general public.
