OpenAI’s newly integrated image generator in ChatGPT has taken off with a bang. However, due to overwhelming usage, Sam Altman has announced temporary limitations to manage the strain on their infrastructure. This measure affects all users as the company works on technical optimizations.
Amazon co-founder MacKenzie Scott has donated over $19 billion to charity in just five years
Diamond batteries powered by nuclear waste promise 28,000 years of clean energy
Servers at Their Limits!
Since the release of the new image generator version, users have been zealously experimenting (guilty as charged!): crafting Ghibli-style illustrations, reimagining famous scenes, and producing stylized portraits. This enthusiasm has far exceeded internal predictions, as Sam Altman shared on X.
In response to this overwhelming demand on hardware resources, OpenAI has decided to temporarily cap the number of creations each user can make, while they work to enhance the system’s efficiency. No specific numbers were disclosed for paid subscribers (though the situation appeared to improve by evening).
Rollout Slowed for Free Users
Already on March 26, Sam Altman had announced that the launch of the image generator for free users would be delayed, also due to server load issues. Despite high expectations, actual usage exceeded anticipated capacities. However, he noted that images in ChatGPT are far more popular than we anticipated (and our expectations were already high)
. When it eventually becomes available to free users, the generator will be restricted to three images per day, as confirmed by OpenAI’s CEO with his new Ghibli-style avatar!
To recap, this update allows unprecedented customization of images. While generating images from simple text was already impressive, this new feature takes it a step further by allowing users to transform an existing photo by applying a specific visual style, create novel illustrations and scenes tailored to user preferences. Users can also add elements or alter details to refine an image.