ChatGPT for educators: Using AI to Create Classroom Content
- chrejsa
- Mar 31
- 3 min read
Updated: Apr 2
Using AI to Teach Color Theory: How ChatGPT-4o’s Image Generator Changed My Workflow
This weekend, the creative and education communities have been buzzing with excitement over the release of OpenAI’s upgraded GPT-4o—particularly the new image generation capabilities. Among the most talked-about improvements are its sharper visual outputs, better prompt understanding, and a long-awaited feature: the ability to display legible text in images.
For educators, this is going to be a game changer.
I experienced this firsthand while finalizing a lecture on color theory—specifically, value, hue, and chroma. I had originally found a great reference image online to demonstrate the concept of hue, but I couldn’t remember where I found it or what the usage restrictions were. Did it require attribution? Was it licensed for classroom use? Could I modify it? Instead of spending more time digging through Google or stock photo libraries, I decided to try something new: I uploaded the image to ChatGPT-4o and asked it to recreate it using crayons—a visual that ties perfectly into how I explain hue to my students (like choosing a crayon from the box).
The first version wasn’t perfect—the crayons were oddly pointed on both ends—but I refined the prompt and got a much better result. A few letters were a bit wobbly, and the red-violet hue leaned a little too red for my taste, but overall? It worked. And it only took me a minute or two in Photoshop to tweak the color and clean up the text. The result was a fresh, custom-made image tailored to my lecture—created without needing to license, attribute, or worry about copyright.


To be clear, I’m not “stealing” an idea here. A 12-color wheel is a standard educational concept used across design, art, and science classrooms. What I created was a personalized teaching tool that visualizes a universal idea in my own way, with visuals that align with how I teach.
ChatGPT for educators will be a useful tool. This is where I see real potential in AI tools like GPT-4o: they’re not here to replace artists or original thinkers. I firmly believe artists are still the only ones who can truly express human emotion, creativity, and individuality. But when it comes to building simple JPEGs for PowerPoints, slide decks, websites, or visual references—especially under tight deadlines—AI image generators can be an incredibly useful assistant.
Of course, there’s a bigger conversation happening here. As educators, how do we credit AI-generated visuals? Are you labeling them clearly in your materials? Is it fair to call it your own work if you heavily modify or build upon them? Does “AI-generated with modifications by [Your Name]” strike the right balance?
I’ll explore that further in an upcoming post—especially as I test GPT-4o’s ability to handle anatomical and scientific illustrations, which are notoriously tricky for AI to get right.
But in the meantime, I’m curious:
Are you using AI-generated images in your classroom?
Do you label them?
How much editing do you do before calling it “yours”?
Where do you draw the line between convenience and authorship?
Let’s talk about it.
Are you using AI-generated images in your classroom?
Yes, as-is.
Yes, but I modify them.
No, AI undermines credibility.
Comments