Google has introduced a new AI tool called Whisk, which allows users to generate images without lengthy prompts. Whisk can combine multiple images that you provide, either through prompts or by uploading images, to create a unique AI-generated image. The tool lets you specify three elements: the theme, the scene, and the style of the image.
To use Whisk, users outside of certain regions may need a VPN. Initially, the tool offers a selection screen where users can upload an image to create something like a stuffed animal. For example, using the t3n logo, Whisk generated a bright red stuffed animal with button eyes. This initial creation can be further refined using the editor, allowing full use of Whisk’s features.
Whisk allows users to adjust the scene and style of the image through prompts. These prompts must be in English, even if the interface appears in another language. For example, a prompt like “sitting in an office” initially generated an image of a businessman at a desk. By replacing the businessman with the stuffed animal in the prompt, a new image was created.
Users can also define the style, such as a Christmas theme with a Santa Claus in a comic style from the 1980s, featuring bright colors and thick outlines. The result was a festive image of the stuffed animal surrounded by cookies, milk, and presents.
Whisk is still in its alpha phase, so some issues are expected. One notable problem is consistency between images. The stuffed animal’s appearance changed significantly within a short time. Additionally, Whisk sometimes struggles with text, especially after multiple prompts. Starting a new creation or reminding the tool of the correct spelling usually resolves this issue. Other common AI issues, like extra limbs or misplaced objects, also occur. For instance, a piece of pizza appeared in a Christmas-themed image.
Despite these challenges, Whisk offers a fun and engaging way to create images, even for users without prior prompt experience. With further development, it is hoped that Whisk will become more accessible to users worldwide without needing workarounds like a VPN.
Google’s Whisk provides a creative platform where users can experiment with image generation and achieve interesting results. As the tool develops, improvements are expected to address current limitations, enhancing the user experience.