Hi Reader,

I'd been training for this moment.

A very long travel day had gotten even longer. The first of our three legs was delayed because, as far as I can tell, the airline forgot where the plane was. It was in "extended parking."

(Between you and me, I don't think that's a thing. 😂)

So we missed the next flight. And the one after that. And the van we'd chartered at the destination.

The kids were champs. They always are. But we'd been up since before 4am to wait for that lost plane, and after a long day of waiting, we had to wait some more.

That's when I found a yo-yo in my backpack.

I pulled it out and put on a show at the shuttle stop.

I'm not very good. Thankfully this crew is easy to impress. I think what they liked was watching me try and miss, wind the string back up, and try again.

I've been thinking about that moment ever since, because it's the same thing I watch leaders struggle with every week.

Where leaders get stuck

I spend my time with the leadership of companies, building AI strategy and readiness. Some are huge, like Dropbox. Most are mid-sized engineering and construction firms you've never heard of.

I see the same thing over and over: the CEO, the COO, the principals, they want to do right by their people. They are rarely confident with AI themselves. They feel behind and out of their depth, and they're embarrassed about it.

So they wait. They plan to set direction once they understand it better, once they've taken the course, and once they can speak to it with authority in front of the team.

Meanwhile, their people are already doing cool things with it.

Earlier this year I ran a full AI discovery inside a US mid-market civil engineering firm, about 250 people across a dozen offices. I interviewed people one-on-one in every office and discipline.

There was no program, no training, and in several cases no company license. Still, roughly one in eight people had already built something real with AI, on their own and without any funding. One proposal coordinator had been paying for a premium subscription out of her own pocket for two years (!!!). Two coordinators in the same office were running parallel systems without knowing the other existed.

Leadership had no idea who these people were. And this team has GREAT leadership.

That's the gap. The people at the top wait until they feel ready, and the people who already have the skills wait for someone to say it's okay.

They need permission. And they need boundaries to build within.

What your team needs from you

Leading well right now starts with accepting that you don't have to be the expert. Your team doesn't need that from you. They have the curiosity, and several of them already have the skill.

They're missing one thing, and it has to come from you: a clear direction that it's safe to use AI at work, which tools are allowed, and that you want to see what they've built.

There are five things only you can give your team:

1. Resources. Licenses on a business account, so people stop using the free version or personal accounts. For a firm that handles client data, proposals, and project files, this is all the more important. No licenses and no guidance = MORE risk.
2. Time. An hour a week for learning. If AI work only happens after hours, it stays siloed.
3. Permission. Tell people plainly that you want them to experiment, so the coordinator paying for her own subscription knows she can show what she's built.
4. Good constraints. Which tools you've approved, which data people keep out of them, and who they show the work to.
5. Humility. Your people want to see you trying, learning, failing. It will inspire them to do the same.

I wrote a few months ago about how the tools work, but people just run out of ideas. In the firm I described, the ideas were already there. They were sitting with the proposal coordinator, the controls lead, and the IT analyst who'd been doing the research at home.

As the leader, you make it safe for people to bring those ideas forward. Then you watch things start to click.

Back at the shuttle stop

That day I did a solid Walk-the-Dog. An overly aggressive Around the World. An okay Baby-in-a-Cradle.

Beyond that, it was just a lot of trying. Winding up the string. Trying again.

That's all it took. The kids didn't need a professional. They needed their dad to show up and give it a go.

You don't have to be the expert. You do have to be willing to show up and try.

And if I can do it, so can you.

I work with teams on exactly this: getting leadership unstuck so the people already building can do it in the open. One hour a week. Your actual workflows. Let's connect.​

Keep building,

-Dan

P.S. Next week I'll share more about Pacemark, the AI maturity model I've been building to help a leadership team see where they stand and what to do next.