‘Please call ASAP’: OpenAI CEO Sam Altman seeks urgent support to handle overwhelming Ghibli trend demand

Almost a week since OpenAI added its new image-generation feature to the GPT-4o model, users have been indulging in the fun of making and recreating images in the classic Ghibli-style art. The new feature has swept the internet, with many eager to see their imagination turned into reality in such a visually striking manner.
OpenAI’s integration of the GPT-4o image-generation feature has shown immense promise, particularly with its ability to generate Ghibli-style artwork that users are raving about. However, the overwhelming demand has led to technical challenges that OpenAI is actively addressing. As the company works to scale its infrastructure and optimize its systems, users can expect some delays and occasional service disruptions. Despite all these glitches, GPT-4o is a very effective tool that has the capability to revolutionise image creation in the near future.

The demand for this feature has been so massive that OpenAI CEO Sam Altman was forced to ask users to dial back the use of the system. Altman admitted that OpenAI is presently experiencing technical challenges as a result of capacity-related issues. This has resulted in delayed updates and service outages. In an X (formerly Twitter) post, he stated that people could expect new releases to be pushed back and cautioned that certain services might break or run slower while the company works through these technical issues.
"We are bringing things under control, but you can expect new releases from OpenAI to be slowed down, things to break, and for service to occasionally be slow as we work through capacity issues," Altman wrote.


To meet the surge in demand, Altman called on the tech community for help, calling for GPUs with over 100,000 concurrent tasks. His call showed the serious pressure that OpenAI servers were facing and their necessity for extra capacity to serve the large number of requests for image generation.
"Working as quickly as we can to actually get things humming; if anyone has GPU capacity in 100k blocks that we can get as soon as possible, please call!" he continued.
The post elicited a deluge of reaction on social media, and the names of several companies were mentioned in a bid to assist with easing the issue. As the demand for the image generation feature skyrocketed, many users voiced their frustrations. Some users reported that the image generator had stopped working effectively, while others noted that the text generation quality had returned to a lower standard. One user expressed dissatisfaction, stating, “The image generator is not working as well now. The text generation is back to being mediocre. I’m a paid subscriber. What’s the deal with that?”


Another commentator proposed that OpenAI could consider releasing features to Pro users with rate limits instead of holding back updates altogether. Some even bashed the move to open up the access to free users, claiming it was one of the causes of the server overload.
Honestly, that move was a little weird. The premium users already have a struggle getting the service to work because of server troubles and really draconian filters blocking even basic prompts. And then free users got access too, just piling on the overloading with all their 'make it look like Studio Ghibli' prompts," one commenter said.

Responding to the tremendous demand, Altman last week declared temporary rate limits, which exposed that the firm's GPUs were being "overloaded." In spite of the challenges, Altman comforted users that OpenAI was actively engaged to optimize the system, enhancing both speed and stability of the image generation service.
“We need sleep too,” he humorously added on Sunday, acknowledging the strain placed on both the team and the infrastructure. He also asked users to slow down with the image generation requests, recognizing that the sheer volume was too much for the servers to handle at once.

Even with these pains of growth, OpenAI is bullish on the prospects of GPT-4o. The new image-generating software is being touted as the firm's "most sophisticated image generator to date." It represents major upgrades over its forebears in terms of accuracy, text rendering, and object relationship comprehension. While earlier models like DALL-E produce images in a single pass, GPT-4o generates them step by step, leading to more realistic and detailed images.

Despite the current setbacks, OpenAI remains committed to improving its image-generation capabilities. As the company works to optimize its systems, users can expect a more seamless experience in the near future. For now, patience is key as GPT-4o continues to evolve, promising even greater advancements in AI-powered creativity.
While delays may be frustrating, the buzz around GPT-4o’s Ghibli-style images shows just how excited people are about AI-powered creativity. As OpenAI fine-tunes its systems, the wait might just be worth it—because the future of image generation isn’t just coming, it’s already drawing itself, frame by magical frame.
Also Read | Are you misinterpreting your dog’s emotions? Study reveals most owners are