I Tested Google’s New AI Tool That Creates Fun Images

Google has introduced a new AI tool named Whisk, designed to streamline the process of image generation. Utilizing Google’s advanced Imagen 3 model, Whisk allows users to create images not just from text prompts but by starting with existing images.

Currently in an experimental phase, Whisk has a user-friendly interface once you get accustomed to it. According to a blog post from Google, the tool is geared toward “quick visual exploration” rather than producing pixel-perfect imagery.

The exploration experience of Whisk feels dynamic compared to traditional text-based tools, which often require precise wording to generate an image.

After you navigate through the introductory pages, which outline key functionalities and provide options for email subscriptions and a privacy policy, you’re taken to the main interface of Whisk. There, you’re greeted with options to choose from various image styles, like a dinosaur plushie or an enamel pin.

Next, users can upload an image to serve as their topic. However, I encountered difficulties uploading a photo of a smartwatch. After switching to a more cartoonish image, the process completed swiftly, leading to the creation of plushie representations of three mythical creatures.

Upon generating an image, you can enter an editing section that includes a text prompt area. By simply using a prompt like “the character is eating ice cream,” I was able to create more images featuring the same creatures enjoying ice cream cones.

If you’re inclined to start fresh, you can select the “start from scratch” option, allowing you to upload your own images or enter text. Adding additional context for your characters is possible from the start, and if you find yourself stuck, an “Inspire Me” button is available to help generate ideas.

The tool features a “My Library” section, where you can access all the images you’ve created. In this area, users have the option to enable or disable the library function if they wish to avoid saving their creations. You can also delete individual items or clear the entire library if needed.

Whisk shares similarities with the Microsoft Designer tool, which allows for the creation of Funko Pop! figures. However, while Microsoft Designer is limited to text prompts, Whisk offers the flexibility of also starting with images.

Despite its image-based foundation, Whisk still allows for the addition of text prompts. Google has included this feature in recognition of the tool’s occasional limitations, ensuring users can refine their inputs whenever necessary.