I think the reason AI feels so hard right now is that you're running two operating systems at the same time.

You've got the way you've always worked — the workflows, the habits, the muscle memory you've built over years. And now you're trying to layer a completely new way of doing things on top of it. Not replacing one with the other. Running both. Simultaneously. While still being expected to deliver at the same pace.

Two systems, one person

Think about what that actually looks like day to day. You've got a deck to build, so you do it the way you've always done it — because you know it works and the deadline is Thursday. But in the back of your mind, you're thinking, "I should probably try doing this with AI." So maybe you spend 30 minutes experimenting, it doesn't quite work, and now you've lost time on both fronts.

You're not learning. You're dabbling. And dabbling can be exhausting.

I see this constantly with clients — leadership teams who want to integrate AI into their marketing, their strategy work, their ops. But they can't pause the business to do it. So the whole organization ends up running the old system and the new system in parallel, and everything just takes longer. The thing that's meant to make you faster is actually making you slower — for a while.

The wrong question

Most leaders I talk to ask some version of this: "How do we carve out time for our teams to learn AI?"

I think that's kind of the wrong framing. It treats AI like a skill to acquire — run a workshop, give people access to tools, check the box. But what you're really asking people to do is change how they work. And that's a transition, not a training program. It needs a plan, it needs sequencing, and it needs someone making deliberate decisions about what to migrate and when.

Because you can't just tell people "stop doing it the old way." That's irresponsible when there are revenue targets and client deadlines on the line. The old OS is proven. It works. You need it to keep running while you bring the new one online.

What I've learned from the other side

I've been through this transition. I run an AI-native practice — I built a product on it, I write with it, I use it across every client engagement. My day job is AI, which means I've already gone through the period where everything felt slower and harder before it got better.

A few things I think are true, having come out the other side:

You have to decommission deliberately. You can't run both systems forever — that's where the burnout comes from. But you also can't just flip a switch. The move is to pick specific workflows, migrate them properly, and let the old version go once the new one is proven. One at a time. Not everything at once.

The wins compound, but they start small. One workflow that clicks. One task that used to take two hours and now takes twenty minutes. Those small wins are what build the confidence to shut down the next piece of the old system. You need to let people experience those before you ask them to change everything.

Someone needs to own the transition. This is the part most companies skip. AI adoption without someone driving it strategically just turns into a bunch of individual experiments that never connect. Someone needs to decide what gets migrated, in what order, and make sure it actually sticks — not just hand out licenses and hope.

The other side is worth it

When the new OS is actually running — when AI isn't an extra thing you're doing but just how you work — it compounds. Fast. The workflows get tighter. The thinking gets sharper because you're spending less time on the stuff that was eating your day. You start to see opportunities you couldn't see before.

The in-between is kind of brutal, honestly. But having been through it, I can tell you the other side is a fundamentally different way of working. And it's more fun. I'll just say that.

The question for most leadership teams isn't whether to make this transition. It's whether you're going to manage it deliberately or let everyone figure it out on their own — and hope it works.