Google announced new AI tool called Whisk which allows you to generate images using other images as a tooltip, rather than requiring a long text tooltip.
With Whisk, you can offer images that suggest what topic, scene, and AI-generated image style you want, and you can show Whisk multiple images for each of these three things. (You can also fill in text prompts if you want.) If you don’t have images handy, you can click the cube icon to have Google fill in some of the images in the responses (though these images also appear to be generated by artificial intelligence). You can also enter text in the text box at the end of the process if you want to add additional details about the image you are looking for, but this is not required.
Whisk will then generate images and a text prompt for each image. You can favorite or download the image if you are ecstatic with the results, or you can refine the image by typing more text in the text box or clicking the image and editing the text prompt.
IN blog entryGoogle emphasizes that Whisk is designed for “fast visual exploration, not pixel-perfect editing.” The company also says that Whisk may “miss the mark,” which is why it allows you to edit basic suggestions.
In the few minutes I used this tool while writing this story, it was fun to tinker with. Generating the images takes a few seconds, which is annoying, and although the images were a little weird, iterating over everything I generated was fun.
Google says Whisk uses the “latest” version of the Imagen 3 image generation model, which she announced today. Google also introduced Veo 2, the next version of its video generation model, which the company says understands the “unique language of cinematography” and is “less likely” than other models to hallucinate things like extra fingers (one of these models is likely OpenAI’s Sora). Veo 2 will be the first to appear in the Google VideoFX program, which can be found on the Google Labs waitlist, and will expand to “other YouTube Shorts products” next year.