Adobe usually saves its ‘Sneaks’ – previews of prototype features that may or may not ship – for its own October MAX conference. But here at Figma’s Config 2023 event in San Francisco, it decided to share one a little early.
The company has been very busy in the AI sphere in 2023, releasing its own generative model, Firefly, back in March. Firefly features have been gradually making their way into the likes of Photoshop and Illustrator – and a new tool, codenamed Gingerbread, could transform text-to-image generation.
Gingerbread is a work-in-progress feature that lets the user, with the help of 3D modelling, adjust the angle and perspective of a text prompt-generated image. “AI tends to put things smack bang in the middle of the image,” said Brooke Hopper, principal designer for emerging design at Adobe. “But what if I want to compose around it?”
this is amazing.@adobe just showcased project ''gingerbread'' at #Config2023 — a image generative AI using 3D objects to tailor your image.Here, Brooke Hopper used a simple cylinder and a text prompt ''whisky glass'' to create this. pic.twitter.com/a1BI3fZ1A5June 22, 2023
Indeed, the fascinating thing about this particular preview is the element of aesthetic control it adds to the process of AI image generation. At the moment, users are limited to the specificity of the text prompts they can write, but Gingerbread hints at a more creative level of customisation, letting users drag and drop a simple 3D shape to determine the appearance of objects in the subsequently generated image.
Time will tell when (or if) Gingerbread will drop, but it’s certainly a tantalising glimpse at the future of text-to-image. In the meantime, speaking of Adobe MAX, at last year’s conference the company had a pertinent message for artists worried about the rise of AI.
