Eric Giesecke is the CEO of Planet DDS.
Most conversations about AI leadership tend to sound alike, with transformation road maps and bold predictions about what the organization will look like in five years. That language sounds impressive, but it skips past the part of the job that keeps me up at night: how we get there without breaking what already works.
The tension few leaders admit is that you have to build the organization you’ll need 18 to 24 months from now while still running the one you have today. You cannot pause operations to retool. Customers still need service and payroll still needs to run. As a result, you really need to do two jobs at once, and most days they pull against each other.
I think of it as renovating a building while people still live in it. You can’t tear out the plumbing and tell everyone to come back in a year. You’re rerouting water while the taps still need to work. That means running temporary lines so nobody loses service and telling residents which faucet to use this week and why.
In an AI transition, the residents are your employees and your customers. The work still has to get done today, even as you change how it gets done tomorrow. The CEO’s job is to sequence that change and protect what’s working while it’s being replaced, while keeping people informed enough that they trust the renovation instead of fear it. It gets harder once you factor in the people doing the work, because they don’t all move at the same pace.
People Move At Different Speeds, And That’s Not A Problem
What surprised me most, even after years of leading organizations through change, is how unevenly adoption spreads across a company. Give the same tool, training and deadline to five departments and you will get five different response curves. Some teams run with it immediately, while others need more time. A few will quietly work around it and hope nobody notices.
The instinct, especially under pressure to show results, is to treat the slow adopters as a compliance issue: Send another email, set a harder deadline, loop in HR. But when someone with real skill and a track record of success is dragging their feet on a new way of working, often it’s because something else is going on: unclear expectations or a legitimate gap in the capabilities of the tool itself. Maybe it’s an unspoken fear, usually about what their job becomes once the tool works.
So I try to get curious before I get directive and ask teams what they see as the main barrier. Sometimes the answer is that the person needs more support to build a skill they don’t have yet. Sometimes they have the skill and the motivation and I just haven’t given them a clear enough reason to change something that already works for them. Those are different problems, and they require different responses. I’ve learned that neither gets solved by a companywide memo. Here is what I recommend leaders do instead.
Say The Quiet Part About Jobs Changing
Some leaders avoid or ignore the hardest conversation, which is what happens to people’s roles during AI tool adoption and transition. I’d rather have that conversation directly than let people fill the silence with worse assumptions. In my own organization, I’ve told people plainly that most of our jobs will look different within a year. I don’t mean they’ll disappear. But the work will shift from producing to reviewing, from doing the task to deciding whether the machine got it right.
That can be an uncomfortable thing to say to a team. It’s also more respectful than pretending it isn’t true. People can handle hard news. What wears them down is watching the ground shift underneath them while leadership insists nothing has changed.
Another part of the job that rarely makes it into strategy decks is deciding, in advance, what requires a human sign-off. Once you start giving AI real authority over workflows, whether that’s scheduling, communications or decisions that touch a customer, you need clear thresholds for when it acts on its own and when a human should intervene to stop and ask.
It’s better to over-define those boundaries early and loosen them as trust builds than to discover deep problems. More than anything, AI transitions are leadership and process decisions that need to be made by someone who understands both the operational stakes and the people affected by them.
Managing The Friction
When I think about what my job looks like during this AI transition (or any other major transition), a lot of it is managing friction day after day, and moving fast while keeping things stable. Some teams will be ready before others, and that’s okay.
We also must accept that there is often a difference between what technology can do and what people are ready to trust it with. A tool can be capable on day one, but adoption happens one workflow, one team, and one good experience at a time. Nobody can hand you a playbook for this because every organization’s friction points are different. The job is finding those and working through them without letting the taps run dry.
Navigating this well has everything to do with showing up week to week and having the uncomfortable conversations early, not worrying about five years down the line. It means treating every person’s resistance as its own puzzle instead of a compliance problem. So I keep checking the taps. Most days they work, and when one doesn’t, I want to hear about it early, and I want whoever tells me to know I’m glad they did.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?