ChatGPT‘s built-in image generation feature is now available to everyone. Last week, OpenAI CEO Sam Altman said that the company was delaying the rollout of the feature to the free tier “for a while” because the tool was much more popular than they expected. But over the weekend, the company made the feature available to free users, allowing them to generate images from ChatGPT without having to switch to OpenAI’s DALL-E generator. Before the free tier was rolled out, the tool was only available to subscribers of the Plus, Pro, and Team plans.
Altman previously said that free users would get a limit of three images per day. Based on our experience, some accounts are limited to only one image per day. Those who manage to create more than one will still have to wait from a few minutes to a few hours. One of the most popular uses we’ve seen so far is to create Ghibli-style images using real photos. Users simply upload the photo they want to use and then give ChatGPT natural language instructions to create a Ghibli version of it. This trend has again raised concerns about the legality of using copyrighted works as training data for artificial intelligence. It also brought to mind the reaction of Hayao Miyazaki, the founder of Studio Ghibli, when he was shown an animation created by artificial intelligence a few years ago. “I am deeply convinced that this is an insult to life itself,” the director said.
In his tweet, Altman said that the tool can still exhibit erratic behavior and deny some generations that should be allowed. OpenAI is still fixing this issue. The company also told TechCrunch that it “takes action” if a generated image violates the company’s rules, although it did not specify what exactly those measures are.
