OpenAI He published a new version of his GPT 4 -based image creation tool yesterday. Which Before it was Dall-e now changed and is simply an extra function within Chatgpt. Faithful to the rapid development times that the generative artificial intelligence models bring, this new image creation tool has enormous ductility, and a shocking ability to generate ultra realistic images. And also of the others.
The initial kick was given by the own Sam Altman, Openai’s CEOwhen presenting this new version of the generation of images within Chatgpt, when it showed how it is possible Take any image and transform it into a drawing in the style of the Ghibli studio films (Pony, My neighbor Totoro, Chihiro’s tripor the recent The boy and the choza). Reach to ask ChatgPP (configured to use its latest version) “Take this image and rebuild Ghibli style”
In the networks, the users quickly understood how this alternative works and began to put it into practice taking historical photos to imagine them as a frame imagined by the great Hayao Miyazaki.
The command It also works in Grok, the generative AI of Xalthough it is not so faithful to the style of the Japanese animation study.
Users found quickly that they can also imitate other visual styles, such as Muppets or Aardvark’s plasticine (from the series of films of films Wallace and Gromit).
The effectiveness of the tool re -discounts the value that the tools of AI put to originality and copyright; Already at the beginning of 2023 several visual artists sued the most popular tools at the time (Stable Diffusion, Midjourney, etc.) found that many users created images with ia copying their style, which can be seen as a compliment, but also as the banalization of their own style, a distinctive way of creating a work that the artists took a long time to find, but that can now be used in a massive way.