OpenAI has decided to withdraw its upgraded image generation feature from the free tier of ChatGPT, just a day after its launch. This update, which utilizes the advanced GPT-4o reasoning model, aimed to enhance realism in AI-generated visuals and text.
The announcement follows an unexpected spike in user interest, particularly in the unique style of the famous Japanese animation studio, Studio Ghibli. The GPT-4o model employs an innovative “autoregressive approach” for image creation.
This method generates images from left to right and top to bottom, contrasting with earlier models that produced images simultaneously. The design aims to improve both the accuracy and lifelike quality of the generated imagery.
Additionally, the new model addresses past shortcomings by creating clearer and more coherent text within images, reducing the prevalence of errors seen in earlier AI outputs. After the feature’s release, users quickly took to social media to share their creations inspired by Studio Ghibli, resulting in a burst of creativity.
Popular reimaginings included scenes from movies like “The Godfather” and “Star Wars,” as well as viral internet memes, all illustrated with Ghibli’s signature aesthetic. Even OpenAI CEO Sam Altman joined the trend, changing his profile picture to a Ghibli-style version of himself.
However, on the same day, Altman announced a pause for the rollout of the image generator for free ChatGPT users, stating that paid subscribers would still retain access. He did not provide a timeline for when the free tier would regain this feature.
The overwhelming popularity of Ghibli-style images prompted OpenAI to reconsider its deployment strategy, highlighting the legal and ethical complexities surrounding AI-generated art and intellectual property rights. The company is taking a cautious approach to assess the unfolding situation and its implications for the future.