Generative artificial intelligences are becoming more popular despite the associated risks

Generative artificial intelligences are on the rise. They are now becoming more democratic in the creation of images or texts despite the associated risks in terms of creating “fake” images or theft of intellectual property. The artificial intelligences of FROM-E 2 and its consumer version, Crayion or Midjourney are able to create photos or illustrations from a simple textual description. As for ChatGPT, it is able to produce text that seems to make sense but is quite soporific.

A French AI hosted at Station F

The latest generative artificial intelligence is French and is called Yokai.AI (formerly Smize). It generates images entirely created thanks to artificial intelligence, with a very realistic rendering. After the launch of the image platform at the end of December, Yokai released its iOS mobile application to generate photos with very specific themes such as a film set for example.

The startup is hosted at Station F. For less than €5, you can generate thematic photos of yourself or your loved ones in Woodstock or in a setting of the TV series Game Of Thrones. With just a few photos, the platform creates a “generator” which then allows artificial intelligence to produce an infinity of new shots of a person in the themes of their choice.

The ethical use of the product is paramount

The algorithm recognizes the face and the human body and generates scenarios that are more real than life. The startup insists that the images are generated in determined themes and situations. This must make any drift or “fake” phenomenon impossible. “Ethical use of the product is a primary criterion,” emphasize the founders of Yokai.AI.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.