
AI has gotten a lot better at creating images.
At first, you had to deal with 6-fingered hands and gobbledy-gook for text, but the results are much improved.
In fact, you can now create images that rival the ones you find from stock photo sites.
The revolution in AI images
But here is what is so special:
- You can get an image that shows EXACTLY what you want it to show
- You can create a series of images showing the same person in different poses, in different places, and with different facial expressions
- You can specify the artistic style precisely
This is revolutionary! I used to spend hours searching on stock photo sites for the right image. No more.
Here’s an example of 2 images I created. The prompt for the first one was:
I need a photorealistic image of a female solopreneur in business casual clothing at a desk that illustrates these words: “You’re trying to grow an online business… …and every step seems to require 10 skills: tech, strategy, copywriting, funnels, offers, pricing, partnerships, messaging, email, setup, tools. And worst of all? You’re doing it all alone. You bounce from video to blog post to podcast, trying to collect answers. But you still don’t know if you’re doing the right thing.” She’s overwhelmed and frustrated, but don’t overdo the emotions.
This is what I got…

Then I asked for an “after” image with the same person. Now, she’s happier because her work is more organized and because she has support from others.

Now I ask you. Could you get these 2 images from a stock image site?
The resolution is probably lower, but you can get exactly what you want. Your images can be perfectly congruent with your message.
That’s powerful.
Using PowerPoint’s Copilot
Yes, you can use Copilot in PowerPoint, if you have a license that includes it. Copilot is still going through a lot of changes but I’ve been able to create excellent images for my slides, using the same technique I described in ChatGPT.
It’s very convenient to create them in PowerPoint if that’s where you’ll use them. Here’s an example…

Just open the Copilot pane and type (or choose), “Create an image for this slide.” Then type in your prompt. My prompt for the last slide shown here is:
African American woman, business casual dress, photorealistic, talking with a client – showing a strong connection between the two
Then I just click the Copy icon at the bottom of the image (it takes a bit of time before it’s complete) and paste it onto my slide or into a placeholder.
Remember that you can always ask for an adjustment if you don’t like the image you get the first time. For example, the first woman is a little wooden and I might ask for a different expression or for her to be facing the “camera.”
Actually, one of the closest alternatives for creating the same “avatar” in a variety of images is PowerPoint’s Cartoon People, because you can create a cartoon person with different facial expressions and positions. But for photorealistic images, AI images give you the best results.
Everyone will use AI differently to create images, but I wanted you to be aware of the possibilities.
Other AI choices for images
I’ve been using the paid version of ChatGPT, so I’m creating my images in the same place I put my regular prompts. I like not having to go to another site.
And as I showed you, I use Copilot in PowerPoint.
Here are some other options:
- Gemini Nana Banana (from Google) can make images inside Google Slides
- Midjourney can create very elaborate, unusual images
- Canva has an AI image generator, but I haven’t used it
Be aware that AI is changing and improving rapidly. Even a few months ago, image creation was filled with mistakes. Now, they are rare and correctable.
Are you using AI for images?
If so, which AI tool are you using? Do you have any tips you can share in the comments?
If not, why not?
Brilliant. Thanks for the insight. Surprising what one can learn if they only pay attention and make the application.