Creativity Cheatsheet: 4 Ways to Create A+ AI Images
Prompt techniques to express yourself with AI
Welcome to another issue of Empowered With AI — the newsletter that helps you use smarter tools to get more done, feel more in control, and take the mental load off your plate.
This week is about creativity and bringing your visual ideas to life with cool new AI tools.
Let’s get into it!
✨ THE EMPOWERED IDEA
My AI Image Obsession
Creativity is a very cool use case for the new generation of AI tools. Basically anyone with ideas can easily bring them to life in new ways.
Fact: I’ve been having way too much fun playing with Text to Image AI creator tools lately. What started as a quick experiment has turned into hours of inspiration research and “just one more prompt” experimentation sessions.
And wow—these tools have leveled up. The photo-realism is so sharp now that sometimes I have to remind myself, this isn’t a stock photo, this was generated in seconds. Even better, the creativity feels endless. Want a neon jungle café with floating lanterns? Done. A serene reading nook that looks like it came straight out of Pinterest? Easy.
Whether your goal is an image for a social post, a birthday party invitation or a concept rendering for a new living design, AI gives creative minds the tool to express themselves.
But here’s the catch.. what you type in matters. And once you know how to craft the right kind of prompt, AI becomes your own on-demand art studio.
Let’s break it down.
🔧 THE PLAYBOOK
The Anatomy of a Good (text to) Image Prompt
Think of an AI image prompt like a recipe: the right ingredients create magic, the wrong ones create… well, nightmare fuel.
Here are the key ingredients you want to mix in:
Subject – Who or what is in the image
Action / Scene – What they’re doing or where they are
Details / Adjectives – Colors, clothing, textures, surroundings
Style – Photography, painting, digital art, cartoon, etc.
Lighting / Mood – Golden hour, moody, cozy, cinematic
Camera Details (optional) – Lens size, angle, realism level
Example:
Instead of typing “woman working at desk” try:
“A confident woman in her 30s working at a sunlit home office desk, sipping coffee, with a modern laptop, Scandinavian decor, soft morning light, photorealistic, shot on a 50mm lens”
See how much richer that is? The AI can now translate your detailed idea because it knows exactly what you’re imagining.
Once you understand the anatomy, the next step is experimenting. Here are four different ways to use AI to generate great prompts:
1. Use the Image Prompt Formula
This is the DIY method: fill in the blanks of a formula and watch the results improve dramatically.
2. Ask ChatGPT to Write the Prompt For You
If you don’t want to think through every detail, let ChatGPT play art director.
3. Reverse Engineer From an Image
Already have a picture you love? Upload it into ChatGPT and let it describe the prompt behind it.
4. Combine Images (Image Remix)
Sometimes one image isn’t enough—you want to blend two into something new.
Note: these techniques are for generating new, creative images - not recreating image of yourself or avatars (character consistency). Those are more complex use cases that will be covered in future issues.
✍🏻 THE PROMPT(S)
1. The Image Prompt Formula
This is the DIY method: fill in the blanks of a formula and watch the results improve dramatically.
Prompt Formula:
[Subject] + [Action or Scene] + [Details / Adjectives] + [Style] + [Lighting & Mood] + [Camera/Rendering Details]
Example:
Instead of “cat in a chair,” write:
“A fluffy orange tabby cat curled on a vintage armchair, soft knitted blanket draped over the side, afternoon sunlight streaming through a rain-speckled window, warm golden tones, cozy atmosphere, ultra-realistic photography style”
Pro Tip: Always include mood/lighting. It’s the difference between “stock photo” and “magazine cover.”
2. Ask ChatGPT to Write the Prompt For You
If you don’t want to think through every detail, let ChatGPT play art director.
Prompt to Copy:
“I want to create an AI-generated image for [purpose].
Here’s what I have in mind: [describe subject, setting, mood, style, colors, and details in plain English].
Please rewrite this into a detailed image prompt for [Midjourney / SORA / Leonardo AI] with professional formatting and vivid description.”
Plain Input:
“A cozy reading nook with a cat by the window.”
ChatGPT Output:
“A cozy reading nook with a fluffy orange tabby cat curled on a vintage armchair, sunlight streaming through a rain-speckled window, a stack of books on a side table, warm golden tones, soft ambient light, ultra-realistic photography style.”
3. Reverse Engineer From an Image
Already have a picture you love? Upload it into ChatGPT and let it describe the prompt behind it.
Prompt to Copy:
“Look at this image and describe it in detail as if you were writing an AI art prompt.
Include subject, setting, colors, mood, lighting, and style so I can recreate it in Midjourney or DALL·E.”
Example:
Upload a photo of a woman reading in a window seat.
ChatGPT Output:
“A young woman in her 20s sitting in a cozy window nook surrounded by pillows and blankets, holding an open book. Warm golden light pours through the window, soft shadows create a peaceful atmosphere, photorealistic style, inviting and calm.”
Now you can swap details to make it your own—like “cool moonlight” instead of “warm golden light.”
4. Combine Images with Sora (Image Remix)
Another creative technique is to blend to inspiration images into something new. That’s where ChatGPT’s Sora’s Remix feature comes in.
How to Do It:
Open Sora in ChatGPT.
Use the Remix option and upload your two images.
Give a clear instruction, like:
“Blend the mountain landscape of Image A with the woman in Image B, placing her naturally in the scene, under golden-hour light.”Sora will generate a fused version that combines both references.
Pro Tips:
Use words like blend, merge, overlay, integrate to tell Sora how to combine them.
Call out mood and style so the final image feels cohesive.
Iterate! Each remix gives you a starting point—tweak the prompt until it matches your vision.
Pitfalls to Avoid
AI image tools are powerful, but a few mistakes can derail your results:
Too vague = weird results. “Dog in park” will look random. Add more details.
Too many adjectives = chaos. Overload the AI and your image turns muddy.
Forgetting lighting. Lighting sets the tone; always include it.
Expecting perfection on the first try. Image prompting is iterative. Tweak one thing at a time until it clicks.
How to Find Inspiration
If you’re stuck, borrow from the world around you:
Pinterest & Instagram: Perfect for mood and aesthetics.
Photography Blogs: Search for “cinematic lighting” or “moody portraits.”
Movie Stills: Think of your favorite film shots and describe them.
Other AI Images: Reverse engineer ones that catch your eye
The more references you study, the easier it becomes to “speak AI’s language.”
Your Move
AI image creation is part science, part play. Once you understand the recipe for a good prompt, you can:
Write them yourself with the formula
Have ChatGPT act as your prompt writer
Reverse engineer existing images for shortcuts and learning
Blend multiple visuals with a Remix
Avoid the common pitfalls, keep your eyes open for inspiration, and remember—it’s supposed to be fun. With just a few words (and maybe a remix or two), you can create stunning, professional-quality visuals that once would have taken hours.
Hope this help!
-Meagan