Google’s Gemini Artificial Intelligence The model came under fire earlier this year for failing to capture human images in a spectacular way. In fact, the people in the images were quite realistic, but it turned out that not only were they historically inaccurate, but secret additions had been made to the commands, resulting in strange images such as the Chinese Viking and the Black Nazi.

In response to the criticism, Google has Twins It tried to limit the ability to create human images. Now google it; Gemini is preparing to offer the human image creation feature in English for advanced, business and enterprise users.

With Imagen 3, Gemini can create human images again

GoogleImage 3 The Gemini issues are expected to be resolved with this upgrade, which is supported by. Imagen 3 is being quietly released through the AI ​​Test Kitchen early this month. This advanced tool “just a few words” of photo quality will be able to create everything from realistic landscapes to watercolor paintings.

GoogleIn February, Gemini stopped displaying images of people after discovering that users were creating historically inaccurate images. Updated Image 3 The model will have built-in security measures. Google states that more successful results are achieved from existing display models.

Gemini offers users photorealistic images of public figures, Content with children or create bloody, violent and sexual scenes. Google plans to open this capability to more users over time.