We’re almost halfway through 2023, and one of the big topics in marketing and technology discussions is the rise of AI tools. These tools have become available in the past year and they’re making a big impact. Learn about Adobe Firefly generative AI.
It’s been a really exciting and also kind of scary time. As a Designer and Illustrator, it’s important for me to keep up with the latest improvements in the field. So a few months ago, I tried a new AI tool called Mid journey.
Before this, I had never used any generative image software like DALL-E 2. These tools can create images using AI, but there have been some issues with copyright. Some people have used pieces of existing images from human artists to generate AI images, which is unethical.
In March, Adobe, a big company that makes design software, launched their own AI model called Adobe Firefly generative AI. Firefly is focused on creating images and texts, and it works alongside their existing software like Photoshop and Illustrator.
There’s a standalone version of Firefly generative AI. with an easy-to-use interface, so people can explore its possibilities. They even have a live show where the Vice President of Design at Adobe presents the concepts behind Firefly and shows some demonstrations of what it can do.
According to the VP, Firefly generative AI. is meant to enhance image creation and manipulation inside Adobe software. It’s meant to make the work of creative designers faster and easier. In fact, they recently added a new feature called “Generative Fill” to Photoshop, which is still in beta testing.
I’ll be sharing more about my experience with Firefly, including the things I found interesting, the good and bad aspects, and what I learned from my first time using it.
Table of Contents
Content Authenticity
One thing that stood out to me about Adobe’s approach to AI tools is their focus on being ethical. They’re really concerned about how AI images are generated. One way they show this is by using licensed Adobe Stock images and copyrighted expired images to train their AI model.
The Vice President of Design at Adobe, Eric Snowden, mentioned that they want to be transparent about what goes into their AI model. They want to create a safe model that people can use for commercial work.
This is important because some other AI tools, like Mid journey and DALLE-2, use proprietary images without the permission of the original creators. Snowden also talked about transparency and a project called Content Authenticity. It’s about knowing the origin of content, who made it, and how it was made. It’s like having nutritional information for content.
With Firefly, all the images created by AI are tagged with content authenticity, so people know they were made by AI. This transparency is important because as AI technology advances, it’s crucial to know how things were created.
Another important aspect is that content authenticity can help identify fake images. When visuals are created with Firefly, they will be labeled as generated by artificial intelligence.
This is important because AI tools are evolving quickly, and sometimes it’s hard to tell if a photo is real or if it has been manipulated. By clearly indicating if AI was involved, it can help prevent the spread of fake news or unethical use of these tools.
The Generative Fill tool itself
I wanted to share something cool about Photoshop with you! There’s a new tool called Generative Fill that lets you do some really amazing things with your pictures. But before I explain, I should mention that Photoshop already has some AI features like “Content Aware” and “Puppet Warp.”
With Generative Fill, you can take a photo and make all sorts of changes to it. I got to try it out myself by downloading the Photoshop Beta version from the Creative Cloud app. Once I had it open, I selected a picture I took during one of my trips.
To use Generative Fill, you need to choose a part of the image that you want to change (you can use the lasso, magic wand, or pen tool). Then, a menu will pop up with the option “Generative Fill.”
I played around with the tool on a picture I had. In the first frame, I had the original picture just the way it was taken. In the second frame, I asked Generative Fill to remove the people and car from the image (and I also made some adjustments with Lightroom).
For the third image, I wanted to add a water puddle with a reflection on the ground. And for the fourth image, I got a bit more creative and asked the AI to put a flying horse in the picture. Now, some of the results were really good, but others turned out a bit strange.
For example, there was a horse with a winged guy on its back, and a floating figure that looked like someone doing a river dance.
It’s pretty amazing what Generative Fill can do, but sometimes the results can be unexpected. It’s a fun tool to experiment with and see what cool changes you can make to your pictures!
Generative Fill great features
Generative Fill has another cool feature: it can add parts to an image that weren’t there before. This is great if you have a nice photo but wish it had a better frame or needed to be wider.
Generative Fill makes photo editing and manipulation much easier and faster. Tasks like removing backgrounds or adding, removing, and replacing elements used to take hours, but now they can be done in seconds if you know how to ask the tool and don’t expect too much from it.
It’s important to remember that these AI tools, including Generative Fill, are constantly improving. The bad results we see now will likely get better in the future.
One possible reason for the mixed results is that the learning database used for training Generative Fill is smaller than a general one, as it’s based on Adobe Stock and expired copyright images. But even so, it’s important to respect copyrighted images and use them in the right way.
Right now, the feedback from the creative community will help improve Generative Fill. In Photoshop Beta, you can give positive or negative feedback for each generated image.
In the meantime, I’ll start using some of the features that can save me a lot of time while still maintaining my own artistic touch. My experience with Generative Fill was like an average “first date” – some things didn’t match perfectly, but there were also some pleasant surprises.
Of course, I’ll need to explore these AI tools more to see where this “relationship” can go. For now, it’s safe to say that Adobe has introduced some game-changing possibilities. Let’s experiment and keep a close eye on what they bring next!
Adobe Firefly AI-powered creative tool
Adobe Firefly generative AI is a new AI-powered creative tool that can be used to generate images, text, and vector graphics. It is currently in beta, but Adobe has released a roadmap of upcoming features that will be added to the tool.
Some of the current features of Adobe Firefly generative AI include:
- Image generation: Firefly can generate images from text descriptions. For example, you could type in “a photorealistic painting of a cat sitting on a windowsill” and Firefly would generate an image of that description.
- Text generation: Firefly generative AI can generate text, such as poems, code, scripts, musical pieces, email, letters, etc. You can provide Firefly with a few prompts and it will generate text based on those prompts.
- Vector-based manipulation: Firefly generative AI can be used to manipulate vector graphics. For example, you could upload a vector file of a logo and Firefly could help you to change the colors, shapes, or other elements of the logo.
Some of the upcoming features of Adobe Firefly generative AI include:
- Text to vector: This feature will allow you to create editable vectors from text prompts. For example, you could type in “a vector drawing of a flower” and Firefly would generate a vector drawing of a flower that you can then edit in Adobe Illustrator or another vector graphics program.
- Text to pattern: This feature will allow you to generate patterns from text prompts. For example, you could type in “a seamless tile pattern of a brick wall” and Firefly would generate a pattern of bricks that you can then use in your designs.
- Text to brush: This feature will allow you to create custom brushes for Adobe Photoshop and other design programs. For example, you could type in “a brush that looks like watercolor paint” and Firefly would generate a brush that you can then use to paint in your designs.
- Sketch to image: This feature will allow you to turn your drawings into full-color images. For example, you could sketch a portrait of a friend and Firefly would generate a full-color image of that portrait.
- Personalized results: This feature will allow you to train Firefly to generate results that are specific to your needs. For example, you could train Firefly to generate images of your product in different scenarios.
Text to image
There’s a tool called text-to-image that lets you make pictures using words. It’s easy to use because it has only two things you can do: type in some words and then click a button to make a picture.
The cool part is that you can also look at what other people have made for ideas. You can see the words they used and even try making a picture based on their words by clicking a button.
Firefly generative AI can make pictures based on the words you type in. You just need to put a description of what you want in the text box and then click a button to make the picture.
Firefly will try its best to understand what you’re asking for and show you four different pictures. Once you have your picture, you’ll have lots of choices and things you can do with it.
Text to image setting
Adobe Firefly has some special features that make it really good. When using other AI art generators, like Mid journey, it’s important to use specific words in your descriptions to get better results. For example, you might mention things like cinematic lighting, close-up photography, or lots of details. But with Firefly, you don’t have to do that. It gives you lots of filters and settings to choose from
Here are some of the settings you can play with:
1. Clear Style: By default, Firefly adds a type of art to your picture. But if you don’t want that, you can click a button to remove it and choose something else instead.
2. Refresh: If you want to try your prompt again and get a different picture, you can click the refresh button.
3. Aspect Ratio: Normally, the pictures are square, but you can change that to other shapes like landscape, portrait, widescreen, or vertical.
4. Content Type: There are four different styles you can choose from: none, photo, graphic, and art.
5. Styles: You can pick different style types to give your pictures different looks. There are 63 options in 7 categories, so you have lots of choices.
6. Color and Tone: If you want a specific color or tone in your picture, you can choose from 7 presets. For example, you can make it black and white, or give it a warm or cool tone.
7. Lighting: You can choose different lighting options to make your picture look different. You can have backlighting, dramatic lighting, golden hour lighting, studio lighting, or low lighting.
8. Composition: If you want your picture to have a certain effect, like a close-up shot or a wide-angle view, Firefly has 7 options to add those lighting effects to your image.
These settings give you a lot of control over how your picture turns out. You can make it look exactly the way you want by choosing different options and filters.
Final thoughts
Adobe Firefly is a new AI tool that is gaining popularity. It’s competing with other strong AI tools like Jasper, Mid journey, and Photosonic. Since Adobe is known for its expertise in digital stuff, Firefly is expected to become a leader in the field soon.
Even though it’s still in the testing phase (called beta), creators can already use it to see what AI can do.
One cool thing about Firefly is that it has features that other AI systems don’t have. For example, you can download images with text effects that have a transparent background.
Also, the pictures you create with Firefly have better quality compared to other AI art generators, which is really great. Another important thing to know is that Adobe takes the responsible use of AI seriously.
They add a watermark and hidden information in each image made with Firefly. This helps people know if an image was created by AI and encourages responsible use of AI-generated images. Overall, Firefly is a powerful tool that shows what AI can do, and Adobe is making sure it’s used in a good and responsible way.
