Adobe updates Firefly AI to streamline tedious Photoshop edits

a firefly on two ai generated hard drives representing adobe fireflyFirefly AI

At its annual Max event, Adobe revealed its latest addition to the Firefly AI tool suite and it looks like another game changer.

Adobe Firefly is getting another update that will allow users to manipulate flat images as independent elements. A new video teasing the latest tool shows a woman with a suitcase gripped in her hand. Firefly can now make the distinction between the woman, the suitcase, and the hands. Once moved, it’s as if the suitcase was never part of the original image to begin with.

Article continues after ad

This is a huge upgrade to the current Firefly tools in the beta build of Photoshop, as well as improving on the already excellent Content-Aware removal tool. As of right now, Generative Fill, the Firefly-enabled removal and replacement tool, can’t figure out everything on screen.

While we’re sure that the newest tool will make mistakes and still require Photoshop skills to fix, it’s streamlining a tedious task.

On top of this, the improved selection tool will be able to distinguish between clothing and also automatically detect unwanted background interferences like people.

Article continues after ad

Adobe Firefly leaves beta but more features are coming

The Firefly toolset is now “out of beta”, but the full toolset is still in active development. Adobe’s web app version still has eight different segments of the software listed under “In Exploration”. Things like text-to-brush and an automatic image extender are still not available.

There are also still concerns over how Adobe is handling Firefly’s learning process. However, updated on September 29, Adobe has stated they’ll compensate all “eligible Adobe Stock contributors”.

Article continues after ad

Last month Adobe also introduced a generative credit system, similar to DALL-E and Midjourney, but it won’t take effect for paid users until November 1.

Adobe’s sprint to bring AI features to its applications has been interesting. It started as a potential marathon with its own machine learning tech, Sensei, and introduced Content Aware Fill in 2015.