CREATIVE GENERATIVE AI
Generative AI is changing creative work.
Generative AI can mimic — but not replace — human intelligence. Learn how it can help you take your creativity to new places.
![Firefly generated astronaut infront of multicolored space background](./media_117a58b5ec18054450465d2836b290681f38edb36.jpeg?width=750&format=jpeg&optimize=medium)
Generative AI is just another tool in your toolbox.
- Like all creative tools, it takes a human to operate generative AI.
- There’s a learning curve to producing quality generative AI creations.
- You can (and probably will want to) edit generative AI imagery in traditional software.
Understand that AI isn’t Terminator or Ultron.
Artificial intelligence has been in creative tools for a while. (Content-Aware Fill is an example of artificial intelligence in Photoshop that photo editors know well.) Generative AI is a more recent development; it’s a technology trained on so much data that it learns to generate new works. Generative AI powers services like the text-only ChatGPT and Adobe Firefly, which has a text-to-image feature.
Generative AI can mimic human intelligence pretty convincingly, but it’s not trying to take over the world. It’s a tool that requires a human to give it direction, usually in the form of a text prompt.
Text-to-image prompt: Midtown Manhattan with cherry blossoms on a rainy day.
Make AI art your own.
Be original.
Play around with text prompts until you learn how to write ones that push the AI in unique directions.
Be picky.
Image generators are good at pumping out a lot of variations. That doesn’t mean they’re all winners. When you train your eye to recognize successful compositions, contrast, and color combinations, you’ll recognize which variations are worth pursuing.
Be empathetic.
If you want a particular kind of person to notice you, you’ll have to do some homework. What’s important to them? Do you understand their cultural context? Bring audience empathy to your creations and you’ll cut through the noise.
Be painstaking.
More traditional software skills aren’t obsolete. Generative AI imagery is usually a starting point, not an end point. You’ll almost certainly want to refine it in software like Photoshop or Illustrator.
![Tallest mountain in the world at night with the northern lights.](./media_1567f955ff66a5fe59dc3a47ff5310ed805e6f129.jpeg?width=750&format=jpeg&optimize=medium)
![Tallest mountain in the world at night with the northern lights.](./media_104a2a7d2cf0f985ee6f1a1cc91fc7b7f69dde4fe.jpeg?width=750&format=jpeg&optimize=medium)
![Tallest mountain in the world at night with the northern lights.](./media_1bcaa4bd1838ae50eecd7165857993e08291a44fd.jpeg?width=750&format=jpeg&optimize=medium)
Walk through the doors AI opens.
Generative AI is an opportunity to branch out creatively. Image generators aren’t limited to flat drawings. Text prompts can produce 3D renderings and even video.
You may discover all sorts of unexpected AI art use cases uses for the technology. Say you want to make a meme that conveys “uncertainty.” To you, a cat nervously balancing on a bicycle is an obvious image, but enter the text prompt in an image generator and see what happens. If the image doesn’t quite match your vision, enter a new text prompt — maybe a nervous horse will do the trick.
One of the coolest parts of generative AI is that you don’t have to stick with your first option. Try modifying your text prompt and tweaking the controls of the AI art generator you’re using until you get something that’s ready to share — or more likely, ready for a final polish by hand.
![Cat nervously balancing on a bicycle](./media_169a92dc2522be35ec05e9c604deaa4acbec97e98.jpeg?width=750&format=jpeg&optimize=medium)
![Horse nervously balancing on bicycle](./media_182b04eb24eb7372a79add4da3d19b3ff8e0735f5.jpeg?width=750&format=jpeg&optimize=medium)
Watch out for AI wrinkles.
As with any new technology, there are things that need ironing out, and generative AI comes with a few big wrinkles.
Not all image generators are created equally — read the fine print of whatever service you use to understand where they source their data from. 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. Other generative AI tools like Stable Diffusion may use a variety of available sources.
Be aware of potential bias. Many image generators allow you to flag a generated image if it shows bias around sex, race, gender, ability, and other traits. Don’t be afraid to flag any images that seem off to you.
Generative AI is still far from perfect, but your willingness to speak up about potential issues and make your voice heard will help the technology develop the right way.
Roll with the changes.
Most of us probably haven’t painted with oils or developed film by hand recently. Instead, we turn to technologies like design software and digital cameras to help us express our ideas. Although getting used to brand-new technology can take time, the results are often worth it.
Generative AI is the next evolution of creative technology for enterprise solutions. It opens up a new world of visual creation to anyone who can write a text prompt — and it offers a huge array of time-saving, imagination-sparking possibilities for people who do creative work every day.
So if you haven’t already, have a little fun with generative AI to see what it can do.
{{you-may-also-like}}