Google has introduced “Whisk,” a groundbreaking artificial intelligence (AI) tool that allows users to create AI-generated images by uploading photos, eliminating the need for input text prompts. This unique feature is designed to provide quick and creative results, an exploratory tool rather than a professional-grade image editor.
Whisk AI enables users to combine subjects, settings, and styles from uploaded images into one balanced, AI-generated image. It also offers the flexibility to remix the final output by modifying inputs and mixing categories, resulting in different styles like plush toys, enamel pins, or stickers. Users can add text to direct specific details, but it is not required to create an image.
According to Thomas Iljic, director of product management at Google Labs, Whisk is designed to allow users to remix a subject, scene, and style in fresh and imaginative ways, offering a quick visual exploration tool rather than a professional-grade image editor. SEO experts and content creators can utilize Whisk AI to bring new visual ideas to life, enriching their content marketing efforts.
Whisk AI, developed by Google, is powered by generative AI technology created by DeepMind, the AI research lab Google acquired in 2014.
Whisk is built on Google’s core AI platform, Gemini, launched in December 2023, and DeepMind’s Imagen 3, its latest text-to-image generator. When users upload images, Gemini generates captions that are then processed by Imagen 3. The tool captures the “essence” of the uploaded content rather than creating a replica, enabling users to experiment with new visual concepts.
However, this can result in differences between the original input and the generated image, such as changes in height, hairstyle, or skin tone.
The technology is still in its early phases of development and is now accessible in the United States through Google Labs. Despite its potential, Google is negotiating generative AI’s drawbacks, including maintaining accuracy and allaying public fears. Google received criticism for creating historically incorrect graphics when it unveiled Gemini’s text-to-image capabilities in 2023.
Competition has been further fueled by OpenAI’s recent introduction of new products, such as the text-to-video generator “Sora.” Analysts see Whisk AI as a calculated step to demonstrate Google’s AI prowess. Wedbush Securities’ Dan Ives referred to Whisk as another “flex the muscles moment” for Google, highlighting the company’s financial commitment to generative AI technologies that will spur innovation in 2025 and beyond.
Source: edition.cnn.com
