The fastest way to make a team afraid of AI is to order them to use it.
I’ve watched it happen. A memo goes out: use AI, multiply yourself this year. The intent is growth, but the reaction is fear. People wonder whether the next hire just got canceled, whether they’re training their own replacement, or how they’re supposed to hit a multiplier nobody explained. The memo says nothing about their expertise, and nothing about how. It’s handing someone a complex piece of software and walking away. The tool arrived. The method didn’t.
I’ve used AI nearly every day since late 2022, and I’ve spent the past few years helping companies, teams, and individuals adopt it: writers, editors, developers, managers, people who were skeptical, and people who were scared. The pitfalls repeat from person to person, and so does the fix. It’s a human approach, and it starts nowhere near the tools.
Start with the expertise, not the AI tool
Enablement that works begins with a worksheet, not a demo. Three columns.
- Name the problem. Write it down the way you’d brief a trusted coworker: what you’re trying to do, what the inputs are, what done looks like.
- Name what you bring. Your expertise, your team’s skills, the assets you already have. This column is usually fuller than people expect, and that’s the point. You were hired for a reason.
- Name the gaps. Maybe you don’t have a coder. Maybe there’s no data scientist on your team, or no time to do the thing you know should be done. Whatever lands in that column is what you hand to the AI. The gap column is the AI’s job description.
That framing changes the relationship. AI stops being a threat to the first two columns and becomes the thing that fills the third. “I know this problem cold, I know what I bring, and I need help with these specific pieces” is a completely different starting point than a blank prompt and a mandate.
Let employees present their own AI wins
Here’s the decision that matters most after a project ships: who presents it.
I could stand up front and demo the team’s wins myself. I do the opposite. When a team finishes a project, the person who did the work presents it to the company. Not me. I’ve seen this moment many times, and it’s the best part of the work: the pride, the excitement of showing off something they built, the coworkers cheering them on.
The reason is simple. When a consultant demos a win, the room thinks: of course, that’s what they’re paid for. When Susan shows the report she automated, the room thinks something else entirely: Susan did that. Maybe I can too. Nobody builds that bridge to a consultant. Everybody builds it to a colleague.
Psychology backs this up. Albert Bandura’s research on self-efficacy identified vicarious experience, watching others succeed, as one of the four sources of belief in your own ability, and he found it works best when the person succeeding is similar to you. A consultant isn’t similar to anyone in that room. Susan is. And Bandura’s strongest source of self-efficacy is mastery, actually doing the thing, which is exactly what Susan got by building the project herself. The model runs on both.
Your role in AI enablement: the audience, not the stage
If you’re leading AI adoption, whether as a consultant, a manager, or the resident enthusiast, your job is to make other people successful and then get out of the frame. The people gain the skills, the confidence, and the credit. You get something better than the spotlight: a team that volunteers for the next project instead of bracing for the next memo.
The same rule applies one level up. If you run the company, you don’t hand down the multiplier. You set the stage Susan presents on.
A mandate gets compliance. A colleague’s win gets volunteers.
This is the model behind my consulting work. If your team wants to get real value out of AI, here’s how we’d work together.
