Sam Altman, CEO of OpenAI, addressed the challenges posed by the recent surge in popularity of the company’s new image-generation feature integrated into ChatGPT. During the New York Times DealBook Summit, Altman noted that the increased usage is causing significant strain on the company's servers, remarking on social media that “our GPUs are melting.”
In response to this demand, OpenAI announced that it would temporarily impose limits on the usage of the image-generation function while efforts are made to enhance system efficiency. The feature, introduced earlier that week, allows users to create high-quality images, which can range from infographics and logos to custom artwork based on user-uploaded photos.
Access to this new capability began with ChatGPT PLUS, Pro, and Team users, and it is set to roll out to users on the free tier of ChatGPT in the coming days, allowing them to generate up to three images daily. Enterprise and educational users will gain access to the feature next week. The introduction of this tool follows the company's earlier success with the Dall-E model, launched in 2021, which was one of the first AI image generators and was later integrated into ChatGPT in 2023.
The new feature has gained attention on various social media platforms, with users sharing their creations, including anime-style renderings. Altman himself adopted an image generated by the tool as his profile picture. As OpenAI continues to refine this technology, the company anticipates significant revenue growth, projecting a tripling to $12.7 billion this year.