Today Adobe have announced Firefly, a family of creative generative AI models coming to Adobe products. In other words, this is Adobe’s answer to A.I image generators like DALLE, Midjourney and Stable Diffusion (learn more about these 3 tools here).
Adobe need to walk a fine line here as their primary market is artists and many artists aren’t especially pleased with these new A.I, especially ones trained off questionable models from an ethics perspective. This means creating tools that assist artists, not replace them, while being trained on an ethical dataset.
Features of Firefly include:
Generative AI made for creators.
With the beta version of the first Firefly model, you can use everyday language to generate extraordinary new content. Looking forward, Firefly has the potential to do much, much more. Explore the possibilities below.
Unlimited creative options, fast.
Think of how many hours you’d save if you could add what’s in your head to your composition instantly. Firefly is gearing up to include context-aware image generation so you can easily experiment and perfect any concept.
Instant creative building blocks.
Imagine generating custom vectors, brushes, and textures from just a few words or even a sketch. We plan to build this into Firefly — plus the ability to edit what you create using the tools you already know and love.
Astounding video edits, simplified.
Change the mood, atmosphere, or even the weather. We’re exploring the potential of text-based video editing with Firefly so you can describe what look you want and instantly change the colors and settings to match.
Distinctive content creation for everyone.
Get ready to create unique posters, banners, social posts, and more with a simple text prompt. With Firefly, the plan is to do this and more — like uploading a mood board to generate totally original, customizable content.
In the future, we hope to enable Firefly to do amazing work with 3D, like turning simple 3D compositions into photorealistic images and quickly creating new styles and variations of 3D objects.
An obvious question is… where does Firefly get it’s data from then?
The current Firefly generative AI model is trained on a dataset of Adobe Stock, along with openly licensed work and public domain content where copyright has expired.
As Firefly evolves, Adobe is exploring ways for creators to be able to train the machine learning model with their own assets so they can generate content that matches their unique style, branding, and design language without the influence of other creators’ content. Adobe will continue to listen to and work with the creative community to address future developments to the Firefly training models.
You can learn more about Adobe Firefly, their new AI system for artists in the video below.