A director at a media company told me something recently that stuck with me: "My 24-year-old assistant has been using AI tools for two years. I tried ChatGPT once for ten minutes. And I'm supposed to be setting our AI strategy?"
This is a new kind of leadership challenge. For most of modern management history, expertise flowed downhill. Senior people knew more because they'd been doing the work longer. They'd seen more situations, built more pattern recognition, accumulated more knowledge.
AI inverts this. The people with the most hands-on experience are often the youngest, the most junior, the ones who've been experimenting on their own time without waiting for permission or strategy documents.
The instinct to fake it
When managers find themselves in this position, there's a natural temptation to bluff. To speak confidently about AI capabilities you haven't actually tested. To set policies based on articles you've skimmed rather than tools you've used. To avoid asking questions that might reveal how little you know.
This instinct is understandable but counterproductive. Your team can tell. And it creates exactly the wrong dynamic for navigating genuine uncertainty together.
A different model: leading the learning
The managers I've seen handle this well do something counterintuitive. They explicitly acknowledge the expertise gap and make learning visible.
This looks like:
Asking genuine questions. "You've been using this more than I have — what have you learned? What works? What doesn't?" Not as a test, but as actual curiosity.
Experimenting in public. Sharing your own attempts with AI, including the clumsy ones. "I tried using Claude for X and it was kind of a mess. Here's what happened." This signals that learning is valued over performing expertise.
Separating AI skill from judgment. Your value as a manager isn't knowing which prompts work best. It's knowing the business context, the strategic priorities, the stakeholder dynamics, the quality bar. Those don't become less important just because you're not the most technically fluent person in the room.
Creating structure for knowledge sharing. If expertise is distributed across your team, your job is to create the conditions for that expertise to become visible and shared. Regular demos, documented workflows, collaborative experimentation.
The ego work
Let's be honest: this requires some ego management. Most of us became managers partly because we were good at the work. Being visibly less skilled than junior team members — at anything — can feel uncomfortable.
But this discomfort is worth examining. What are you actually afraid of? Losing authority? Looking foolish? Being seen as out of touch?
The paradox is that pretending to know more than you do undermines authority far more than honest curiosity. People respect leaders who can say "I don't know yet, help me understand" far more than those who bluff through uncertainty.
The questions that still need you
Even if your team knows more about the tools themselves, there are questions that require your experience and judgment:
- How do we think about quality and standards when AI is involved?
- What are the risks we need to manage — reputational, legal, ethical?
- How does this fit with our broader strategy and priorities?
- What does this mean for how we structure work and develop people?
- How do we move from individual experiments to shared capability?
These are leadership questions, not technical ones. And they're where your experience matters most.
The temporary nature of the gap
Here's the good news: this particular awkward moment won't last forever. You'll build your own fluency. Norms will develop. The gap between "people who've experimented with AI" and "people who haven't" will become less stark.
But how you handle this transition matters. The managers who lean into the discomfort, who model curiosity over performance, who create space for distributed expertise to flourish — they'll emerge with stronger teams and better judgment.
The managers who fake it will just be behind.