Whisk, Google’s generative AI image creation tool, is officially available for Filipinos to tinker with. Launched late last year, Google’s latest experiment explores generative AI and lets you have a more accessible experience to AI image generation.
Instead of typing long and detailed prompts, Whisk will now let you use image prompts. You will now be able create AI-generated images by simply dragging images.
Generative AI uses deep-learning models to make content based on existing data. Certain elements can get lost in text prompts, but Whisk’s can customize how the image will come out with specific image inputs for the main subject, scene, and preferred art styles.
Behind the scenes, Google’s Gemini model automatically writes a detailed caption of chosen images. These are then fed to their latest image generation model, Imagen 3. They claim that this process captures the subjects “essence,” but doesn’t create an exact replica. This allows you to remix subjects, scenes, and styles.
They also highlight that Whisk extracts only a few key characters from images, which makes generated content that may differ from your expectations. Fortunately, it allows you to view and edit underlying prompt at any time to get the output you want.
