A few weeks ago I attended the online version of Adobe Max, the yearly graphics & photography design conference held by the software giant. Coming from a digital marketing background, Adobe Max is my favorite conference in the marketing and graphic design industry. This year's focus was almost exclusively on the introduction of Adobe Firefly - the AI software now integrated into most of Adobe's products, including Photoshop, Illustrator, and Express.

Over the past few weeks I've been experimenting with Firefly in Adobe Express - Express is my go-to product when creating original graphics for social media, as well as when I need to create a banner image for my posts on sites like Publish0x. I was hoping that the introduction of Firefly would speed up the already quick creation process on Express...and it mostly works. But in my experience, Firefly has a ways to go in certain categories before it can be completely relied upon for content creation.

Using Firefly in Adobe Express

To begin with, how exactly does a content creator use Firefly in Adobe Express? Adobe Express has launched a new interface this past month that now incorporates a Generative AI tab on the main navigational menu. To start your image generation, click on this tab and scroll down to access one of four prompts: Text to Image, Generative Fill, Text to Template, and Text Effects. So far, I've utilized the Text to Image and Text to Template prompts to varying degrees of success.

For all prompts, in general, the more text (information) you provide, the better the results. If you're looking for examples or inspiration, each prompt line in the Generative AI section has a handful of examples that you can hover over and see the prompt that was provided. In some of the examples I'll provide below, you'll see that the content generated was probably hindered by the lack of information that I provided as compared to Adobe's  examples. However, I did find that certain topics seem to be out of the AI's comprehension at the moment, too.

Examples of Using Firefly in Adobe Express

Example 1: Wombat Wallet Blog Header Image

My first attempt using Firefly in Adobe Express actually went really well. I wrote my recent article reviewing the Wombat Wallet app, and wanted to quickly create an original banner image for my post. My prompt for Text to Image was "a wombat wallet in purple and orange," and it created this cute little wombat wallet graphic that I used along with Wombat's logo and some text. All in all, it only took me a few minutes to create my banner - no Illustrator required.

Example 2: Blog Header Image for this Article

The blog header image for this article was also created with Firefly in Adobe Express using the Text to Template prompt. To be honest, I find that the Text to Template prompt creates really simple imagery - not quite the same level of quality that I'm used to finding by simply browsing the templates included in Adobe Express. Still, I selected the template I liked best, changed some of the text, sizing, and colors, then saved it for use today. As you can see it's not a bad image - just kind of boring in my opinion.

Original Template Suggestion:

Template After Edits:

Example 3: Cryptocurrency Airdrop

Since I write about potential testnet airdrops rather frequently, I thought about trying to create an original banner image for future use using the Text to Image prompt with the text "cryptocurrency airdrop." In this example, I think the results are a combined failure of both my lack of useful information and the AI not understanding what an airdrop is. As you can see in this group of four image suggestions, some of the images are simply combining the bitcoin logo with...wait for it...air fans that look like they have been dropped. Oof.

Example 4: Space Plasma Physicist

If you've read this far, you're probably doing a double-take at this example's topic. Long story short, my husband is a space physicist, so I tried a few prompts at his suggestion based on his work. Unfortunately I forgot to save some of my early results, but they were pretty bad - a few weeks ago, using the term "physicist" was interpreted as "physician" by the AI, and all of our results included doctors with stethoscopes in the imagery, rather than a more generic scientist. As of this week that term seems to be fixed in the AI generator, but some of the images are still somewhat odd. My latest prompt was "space physicist giving a lecture," - and you can see that one of the options shows an astronaut giving a presentation. But, one of the images actually came out pretty well - so it's been saved in case my husband ever wants to use it on his own social media. :-)

Example 5: Cryptocurrency Learn & Earn

This last example is another where I think the combination of my lack of useful text combined with the AI's lack of resources to produce anything coherent at all. Since I also frequently write about Learn & Earn opportunities when they arise, I used the term 'cryptocurrency learn and earn course' and got these results. I ran into similar images when playing around with some of my husband's more technically detailed prompt suggestions, too: results with absolute gibberish. I can understand my lack of description would lead to these really generic looking signs - but that doesn't explain why everything would be misspelled.

Summary of Using Firefly in Adobe Express

All-in-all, my usage of Firefly in Adobe Express has been a bit of a mixed bag. I've definitely had some cases where it knocked it out of the park on the first attempt; then there have been other times where it seems my topics don't seem to resonate with the AI software yet. But as the AI learns from feedback from Adobe users, I'm excited to see how the software improves and the time that can be saved on digital graphic generation.

Like this article? Find me on...

Medium: @Valdyr | Twitter: @Valdyr | Torum: @Valdyr | Publish0x: @Valdyr | Article: @Valdyr | Phaver: @Valdyr | Bulb: @Valdyr

Valdyr is a "crypto-curious" blogger living in Texas with her husband and their ancient dachshund, Henry. When she's not working as an audio/visual technician or stagehand, you'll frequently find her traveling, hiking or enjoying a good craft beer. For more reading, check out www.laurahofford.com/blog/