Adobe has unleashed a groundbreaking feature in Photoshop, enabling users to generate images from scratch using text-to-image technology powered by Firefly. This innovative tool, first teased in April, allows users to create an image by typing a description into Photoshop’s updated generative AI tool.
I experimented with the feature, inputting the text “Dramatic low angle view of a steamship from the 1800s in a storm with large waves and lightning” in various styles, including anime, watercolor, sketch, and realistic. The results were decent, with the ability to fix any errors directly within Photoshop.
Previously, Photoshop’s Generative Fill feature only allowed users to add, extend, or remove specific parts of an image. Now, users can create images from scratch and then refine them later. According to Adobe’s Erin Boyce, this feature significantly accelerates the creative process, making it easier to bring ideas to life.
The Firefly Image 3 model powers this feature, which has been at the center of a recent artist backlash against Adobe. Creators were concerned that Adobe’s recent terms of service allowed the company to use their work to train its generative AI models without permission.
However, Adobe has since clarified its stance, committing to a “creator-friendly AI” approach that ensures it will never train its models on customer content. The company has also outlined its AI ethics principles, which prioritize accountability, responsibility, and transparency.
In addition to image generation, Adobe has introduced several other features in Photoshop, including an “Enhance Detail” feature in Generative Fill, which refines the quality of generated details. Illustrator users can now take advantage of Generative Shape Fill, Enhanced Text to Pattern, and Style Reference, as well as a new Mockup tool for creating high-quality visual prototypes.
