Google Launches Whisk - A Unique Image Generator
Google introduces Whisk, an AI image generator that creates new images based on uploaded photos instead of text. Could this change how we work with creative creation?
Google has introduced Whisk, an experimental image generator that uses images as prompts instead of text. Whisk is powered by Google's AI technologies Gemini and Imagen, where Gemini automatically converts uploaded images into detailed text descriptions used to generate new images.
According to Google, here's how Whisk works:
Automatic description: Gemini automatically creates detailed text descriptions based on uploaded images.
Generative remixing: Imagen 3 uses these descriptions to generate new images inspired by the original image.
Editable prompts: Users can adjust the generated text descriptions to refine the result.
Focus on Inspiration Rather Than Perfection
Whisk doesn't produce exact copies of original images, but rather what Google describes as the "essence of the image". Results can vary considerably in tone, mood, and scale, making the tool ideal for creative ideation rather than photorealistic artwork.
Google positions Whisk as a tool for "rapid visual exploration", making it possible for creators to quickly test and iterate on different ideas. It's faster and more flexible than many other AI image generators but isn't optimized for pixel-perfect image creation.
Availability and Future
Whisk is currently available through Google Labs as an experimental project. Users can test the tool and give feedback to Google, but there's no information on when or if it will be released to the general public.
A New Era for AI Prompts?
Whisk could pave the way for a new generation of AI tools where image-based prompts simplify creation for users who don't have the time or knowledge to craft advanced text prompts. The question remains, though - what happens when AI-generated images start being used to generate new AI images?