Last month, I was leading a breakout session at Serious Business, a salon industry conference. Standing room only — which already told me something.
I started with a question.
“How many of you have transformation photos you can’t post? The client’s hair looks incredible, the work is some of your best, but there’s hair on the floor, electrical cables running through the frame, and clutter in the background. Something that makes it unusable.”
Every hand in the room went up.
“How many of you have hundreds of photos like that?”
Most hands stayed up.
“Thousands?”
A few people laughed. Still hands up.
This is the photo graveyard. Every service business has one.
What Happens in the Graveyard
Think about what that actually represents. A stylist spends two hours on a transformation. The client is thrilled. The result is genuinely stunning. Someone pulls out a phone to capture it — and then later, sorting through the camera roll, sees the hair sweep that didn’t happen yet, the charging cable draped across the mirror, the cart of supplies in the corner of the frame.
And the photo goes into the graveyard. Never posted. Never shown to a potential client. Never added to the portfolio. Just… sitting there.
Multiply that by every stylist in a salon, every week, for years. One person at my session said she had over 500 photos like that. Another estimated thousands.
That’s not just a missed social media opportunity. That’s evidence of real work that nobody ever sees.
The 20-Second Demo
After the show of hands, I asked if anyone wanted to try something live.
We took one of those photos — a real one, pulled from someone’s phone in the room. Uploaded it to Gemini. I gave it a simple prompt: remove the background clutter, clean up the cables, clear the floor.
Twenty seconds.
The cables were gone. The hair on the floor was gone. The background looked like the photo had been taken in a clean, properly staged environment.
The room went quiet. Not the polite quiet of an audience following along — the quiet that happens when something genuinely surprises people.
Then someone said, “I have 500 of those.”
What had been a graveyard for months became a content library in under a minute.
This Isn’t a Salon Problem
I want to be clear about something: the salon context made this demo vivid. But the underlying problem is everywhere.
Personal trainers take photos in gyms that have other clients in the background, equipment mid-use, awkward lighting from windows. Contractors photograph finished work with construction debris still in frame. Restaurant owners snap photos of new dishes with the setup clutter not yet cleared. Event photographers capture beautiful moments with venue chaos behind them.
Anyone who takes photos of real work, in real environments, has a version of this problem. The work is good. The background isn’t ready. The photo goes unused.
The solution is the same in every case: a few seconds in an AI image tool, a simple prompt, and a photo that was previously unusable becomes shareable.
Why This Demo Works So Well
After three standing-room-only sessions at that conference, I had time to think about why the photo demo landed so consistently.
It’s not because it’s the most powerful AI use case. It isn’t. It’s because it solved a problem people already had, using work they had already done.
Most AI demos show something futuristic. A workflow that sounds impressive but requires you to build something new, change a process, or learn a new tool from scratch. The gap between “I could see how that would help” and actually using it is wide.
The photo demo has almost no gap. You have the photos already. You open Gemini. You upload. You get the result. The only thing that changed is now you know this is possible.
That’s the pattern I keep seeing with AI adoption in non-technical industries. The tools don’t need to be complicated. They need to be specific. A salon owner doesn’t need to know what a language model is — she needs to know she can fix that photo from March that’s been sitting unused.
What to Actually Do With This
If you have a graveyard of photos like this — from any industry, any context — here’s the workflow:
First, sort through your camera roll and pull out the photos that show good work but have problematic backgrounds. Don’t overthink it. If you’ve been putting off posting something because of the background, that’s the one.
Then open Gemini (free to use, no setup required) and upload the photo. Ask it to remove specific elements — cables, clutter on the floor, items in the background. Be specific about what’s in the way.
Review the result. Gemini is good at this, not perfect. Some photos will need a second pass or a more specific prompt. But most of the time, 20 seconds is accurate.
Start posting the ones that come out well. Build the habit. Go back to the graveyard periodically and work through more of them.
The goal isn’t to use AI for every photo. It’s to stop letting great work sit unused because the background wasn’t clean.
The Bigger Pattern
What struck me most about those sessions was the hands going up before the demo.
Everyone already knew they had this problem. Nobody had thought of AI as the solution, because nobody had shown them a solution that worked for their specific situation. Once they saw it — really saw it, live, on a real photo — the barrier disappeared.
That’s what I think is still underestimated about where we are with AI adoption. The tools have gotten genuinely good at specific, practical tasks. The gap isn’t capability. It’s translation — helping people see where the tool connects to the work they’re already doing.
The graveyard of unpublishable photos is just one example. But it’s a good one, because the before is so clear and the after is so immediate.
Check your camera roll. You probably have a version of this.
Thanh Pham is the founder of Asian Efficiency. He teaches AI fluency through workshops and the 4-Day AI Sprint — designed to take people from occasional AI use to real, running workflows.
