Your best people are the ones refusing to use AI. That's not a coincidence.

Half your team is using AI. The other half won't touch it. Nobody's fighting about it out loud — but the resentment is there, and it's starting to show up in the work. Here's why the split isn't actually about AI, and the one move that turns your most resistant veterans into the authority instead of the opposition.

Steve Buckner

Cloud Systems Engineer · MCT · PMP · Azure Solutions Architect Expert. 40+ years in IT and operations. Builder of the Blair AI Rollout Framework.

Published July 2026

🎙 Listen to the Episode

Your Best People Are the Ones Refusing to Use AI. That's Not a Coincidence. Blair AI Rollout Podcast · Season 3, Episode 9 · Steve Buckner

Ray runs plant operations at a mid-sized manufacturer. His younger supervisors cut shift reporting time in half using AI. His twenty-year floor leads won't touch it. Now he's running one plant on two different standards — and the tension is building.

Listen on Spotify

Nobody's fighting about it. That's what makes it dangerous.

Ray runs plant operations at a mid-sized manufacturer. Two of his shift supervisors figured out how to cut their reporting time in half using AI — and they're doing good work with it. His most experienced floor leads, guys with twenty-plus years on that floor, want nothing to do with it.

There's no open conflict. No confrontation in a meeting. But the resentment is there, and Ray can feel it. He's now getting two different qualities of work out of the same plant, and he can't run an operation that way.

His instinct is that he shouldn't force anyone. That instinct is right. But the situation is different from how most leaders describe it — and getting the diagnosis right is what makes the rest solvable.

1. The split isn't about AI. It's about respect.

Consider what this looks like from where the floor leads are standing. Someone has been running that floor for twenty years. He knows which machines run hot in the summer. He knows which operator to put on which line on a Monday. He knows things that were never written down anywhere, because he learned them the hard way over two decades.

And now somebody half his age is producing a shift report in a fraction of the time using a tool he's never touched.

What he hears — even if nobody said it — is that the thing he's best at just got cheaper.

That's not stubbornness. That's a completely rational response to feeling like your expertise is being quietly devalued. Which means this isn't a technology adoption problem. It's a respect problem. And respect problems don't get solved with training sessions.

2. You don't have two standards. You have none.

Ray said he can't run a plant on two standards. He's right about that. But look closely at what he actually has.

His supervisors using AI are producing fast, clean, consistent reports. His floor leads are producing reports the way they always have — which, in Ray's own words, is late and uneven and depends on who wrote them.

Neither of those is a standard. One is a tool being used well. The other is habit.

That distinction matters, because it changes what problem you're actually solving. You don't need everyone to feel the same way about AI — you're never going to get that, and you don't need it. What you need is one clear definition of what good work looks like.

Standard first. Tools second. Not the other way around.

3. Don't mandate. Make the veterans the authority.

Here's where most operations leaders get this exactly backwards. The instinct is to take the fast, clean reports the AI users are producing and declare them the new bar — everyone match this.

Don't do that.

The moment you hold up the AI output as the standard, you've told your floor leads that the tool won and their judgment lost. You'll get compliance if you push hard enough. You will not get buy-in. And on a plant floor, the difference between those two things shows up fast.

Do the opposite. Put your most experienced people in charge of defining what good work actually looks like.

They're the right people for it, and not as a consolation prize. They genuinely know better than anyone what information matters at handoff, what gets missed, what causes problems on the next shift. That knowledge has never been written down anywhere. This is your chance to capture it.

Then bring it back to the whole team: here's our standard, it came from the people who've been doing this longest, and how you hit it is up to you.

This week's AI Hot Tip.

Three things to take with you.

• The split isn't about AI. It's about what adopting it seems to say about the people who've been doing the work the longest. That's a respect problem, not a technology problem — and it doesn't get solved with training.
• You don't have two standards. You have none. You don't need everyone to agree about AI. You need one clear definition of what good work looks like. Standard first, tools second.
• Don't hold up the AI output as the bar. Put your most experienced people in charge of defining it. Now it's not old versus new — it's everyone meeting a standard the veterans wrote.

Related resources.

Common questions.