Whisk Google's new artificial intelligence to create images that will amaze you
Google has a new AI system that will make creating images easier. The company has launched Whisk, a new model that does away with the featured prompts and simply uses other images for clues about what to create.
Whisk works in a very simple way. To create an image, the user only has to add an image that acts as the subject (i.e. the main subject of the image); be it an animal, a person, an object, or any other element. It is also necessary to add another image that acts as a background, for example, a landscape, as well as a third image that acts as a source of inspiration for the style. In this case, it is possible to include an anime scene, for example, so that the image is created with the same aesthetic.
Google’s new AI also lets you enter a very short text description so that the image is more accurate than what the person wanted to create. For example, if a user added a picture of a dog, a space background, and a cartoon style, they can simply put in the description.
Google uses Gemini, its powerful language model, to get a text description of the attached images and send it to Image 3, the company's image generation model, which is responsible for creating the design.
The company also explains that Whisk is designed to be used as a “creative tool” rather than a traditional photo editor. “We built it for quick visual exploration, not pixel-perfect edits. It’s about exploring ideas in new and innovative ways, allowing you to work through dozens of options and download the one you like best.”
Google's new AI that lets you create images from other photos is now available in the US. Users can try it out on the company's Labs website . At the moment, it's not known when it will be available in other countries.