ChatGPT transforms photos into Ghibli style: an impressive feat that raises questions 🎨

Cliquez ici pour lire en français

For the past few days, social media has been flooded with images transformed into drawings inspired by Studio Ghibli. This trend is linked to the latest version of OpenAI’s image generator, which can convert any photo into an illustration faithful to the aesthetic of Hayao Miyazaki’s films.

A technology that is increasingly precise 🎨

This image generator, the successor to DALL-E 3, is based on the GPT-4o model. It can preserve all the details of an original image, including facial expressions. OpenAI has also improved text and infographic handling. One of the major advancements lies in the AI’s ability to respect an image’s context, ensuring coherent variations.

On Tuesday, March 25, American engineer Grant Slatton shared a family photo transformed into a Ghibli-style version. His post generated over 20 million views on X, sparking a wave of experimentation. Since then, famous memes, celebrity portraits, and movie scenes have been « Ghibli-fied. » OpenAI has even integrated this technology into its video generator, Sora, allowing complete animations to be recreated in this style. Now, it’s hard to scroll through your feed without coming across a transformed image.

How does it work? 🤔

Here are the steps to transform a photo into Ghibli style with ChatGPT :

• Go to chatgpt.com or the mobile app and log into your OpenAI account.
• In the options under the text field, click on the three dots, then select « Create an image. »
• Upload your photo via the dedicated button.
• Enter a prompt such as: « Restyle this image in Studio Ghibli style, keeping all details. »
• ChatGPT will apply the Ghibli style to your image.
• Once the image is generated, you can download and use it according to OpenAI’s terms.

A success that raises questions 🎨

While this tool is impressive, it also draws criticism. Some argue that it is a « soulless » version of the Ghibli style, limited to a few characteristic features like large eyes and fine outlines. Hayao Miyazaki himself has always opposed artificial intelligence in animation. He sees this approach as contrary to his philosophy of meticulous and handcrafted drawing.

From a legal perspective, the issue of intellectual property also arises. OpenAI claims to have trained its model with publicly accessible images and a partnership with Shutterstock. However, the legality of these practices remains uncertain, especially in the United States, where several lawsuits are ongoing against OpenAI for using copyrighted content.

Towards restricted access? 🎨

Given the scale of the phenomenon, OpenAI has announced a temporary limitation on the generator’s use, notably capping free users at three images per day. Despite these restrictions, enthusiasm remains high, reigniting the debate on AI’s impact on artistic creation.

We want your opinion!!!
What do you think about this trend? Let us know in the comments ☺️