09/08/2026
If you’re lucky enough to have AI create a character or avatar that you really like, save it. Better yet, create a folder specifically for your AI-generated characters and start building a reference set.
I learned pretty quickly that asking AI to “make the same character again” doesn’t necessarily mean you’re getting the same character again. There’s randomness involved in image generation. You might get something close, but the face changes, the proportions change, clothing changes, little details disappear, and before long the character you started with has a distant cousin showing up in your next post.
We ran into exactly that while developing some of the visual storytelling for Photography Fit 2 Print. John and AI started becoming recurring characters, and once I liked who they were, I didn’t want a different version of them every time I had another idea.
So we changed the process.
Instead of generating new characters for every scene, we created a character set, saved the references we wanted to preserve, and established some rules for how those characters should be used. Now when we want them somewhere different, the idea is to modify the characters we already established rather than start over and hope the generator gives us the same people again.
There’s more going on behind that process than I’m going to put in a Facebook post, but the important lesson is simple: when AI gives you something worth keeping, stop treating it like a disposable generation. Start treating it like an asset.
That has allowed us to take John and AI from one environment to another while maintaining enough visual continuity that they can actually become part of the Photography Fit 2 Print storytelling.
AI can generate an image in seconds. Building something recognizable enough to use again and again takes a little more thought.
And that, to me, is where this stuff starts getting interesting.
Fit 2 Capture. Fit 2 Curate. Fit 2 Print.
