AI image created with text description

Adobe Photoshop turn written descriptions into AI artwork

Generative Fill is a tool that is available in the web version of Photoshop. Photoshop, the popular image-editing program developed by Adobe, already has an active tool for generating photos or images from a short text description.

Adobe Firefly is the name of the artificial intelligence found in the program with the name Generative Fill, a feature that allows you to create designs in a similar way to how we would do it with DALL-E or Midjourney, although in this case it goes further and it is also capable of removing or adding elements in a photograph.

How to use AI in Photoshop

It is important to note that this new tool is available in the Photoshop desktop application, and for now, it is in the first version, that is, its Beta phase. To use it we must:

Download the Photoshop trial app, access Creative Cloud, and click on the “Applications” tab.

In the side menu, click on the section called “Applications in beta version” and install Photoshop.

It may interest you: Photoshop includes artificial intelligence and more than 100 languages ​​to edit

Once the beta version of Photoshop is installed, it is possible to access the program and open a new image from the menu bar under File > Open, or by pressing Command + O.

It is necessary to choose one of the different selection tools that Photoshop offers and highlight the area of ​​the image that you want to modify.

Photoshop will display a new context bar with shortcuts to tools related to image selection, including Generative Fill. It is only necessary to click on that button and write what we want the AI ​​to generate.

The text must be in English and once written it is necessary to click on “Generate” and in a matter of seconds we will have three different examples of designs to choose from. In addition to creating images using artificial intelligence, Generative Fill also allows you to enlarge the background of a photograph with a very high level of realism.

To do this, simply expand the image in the menu bar, click on Image > Canvas size, and select the desired size. Photoshop, by default, will add a blank canvas next to the image.

Finally, the selection tool is chosen and the entire blank area is drawn. The contextual bar will show the Generative Fill button. You just have to click on it and, finally, click on ‘Generate’.

One of the great features of this tool is that it renders designs on a new “generative layer” so that users can easily remove or modify them without altering the rest of the image. While Generative Fill has great potential ahead of it, for now, Photoshop does not allow you to generate copyrighted, branded designs and therefore must be generic.

Other AI-powered features in Photoshop

Adobe has announced several features to improve the editing experience in Photoshop, such as a new preset mode that allows you to view and apply filters to an image in a much more intuitive way without having to perform actions such as adjusting the tones.

At the moment, Photoshop comes with 32 presets that can be easily applied and undone, as well as edited if necessary.

Another of these AI-powered features is Remove, capable of removing elements from an image simply by brushing over them, which is especially effective with large objects.


In conclusion, turning written descriptions into AI art is really cool and opens up a ton of new creative possibilities. As technology keeps advancing, AI artwork is only going to get more insanely realistic and detailed – the future possibilities are mind-blowing! Even someone like me with zero artistic skills can play around with user-friendly tools like DALL-E 2 or Midjourney and now Adobe Firefly to bring my wacky visual ideas to life. It obviously takes some trial and error to come up with text prompts that work, but overall it’s pretty easy for anyone to generate beautiful, one-of-a-kind AI artworks. I can’t wait to see what amazing AI artwork people come up with next!


What are the best apps for AI art?

For most people, DALL-E 2, Midjourney, and StarryAI are awesome to start with. They make it super simple.

How much money do I need to spend?

You can typically get some free credits to try it out first. After that, it’s pretty affordable, like $10-$30 monthly. DALL-E charges just 1 cent per image which is awesome!

What file types does it output?

Usually common image types like JPG or PNG. Some also let you download them as SVG files for expansion later.

Scroll to Top
Top High-Income Skills to Learn in 2023 Create Stunning Images with ChatGPT and OpenAI’s Dall-E 3 7 Step to Become a freelancer in 2023 4 Best freelancing skills for Beginners in 2023 5 Best in-demand Freelancing skills for students in 2023 9 surprising ROI Facts to Maximize Your Returns The 5 Big Ideas of AI: A Comprehensive Guide Apple unveils iPhone 15 Pro and Pro Max with titanium bodies The Evolution of iPhone Prices: iPhone 15 vs. Previous Models YouTube Announces AI-Powered Creative Guidance In Google Ads