Leadership In The Age Of AI: The Work Is Good. The Muscle Isn’t Built
A junior on the team sends through a strategy document.
It’s good. Not “good for a junior” good. Actually good. The kind of document that two years ago would have come from someone with a decade of experience behind it.
And the leader sits with it for a while before responding. Because they don’t quite know what to do.
This is the question almost nobody in leadership development is talking about, and it isn’t the one people assume. It isn’t did they use AI. Everyone uses AI. The question is what you do with the work, and with the person, when the old shortcuts for assessing both no longer hold.
The shortcut that just broke
For as long as anyone has managed people, the work has been a proxy for the person.
You read someone’s output and you inferred a lot from it. Depth of thinking. Judgment under pressure. How much they could be trusted with next. The quality of what landed on your desk told you, roughly, what the person was capable of. It was never perfect, but it was reliable enough to build careers on.
That proxy has come apart. Output and underlying skill have been quietly decoupled. A person can now produce something genuinely sophisticated without yet possessing the experience that the work seems to signal. The document looks like ten years of judgment. The person has been doing the job for ten months.
So the leader is left holding a piece of work that no longer tells them what it used to tell them. And most have no framework for what to do next.
Two ways to get it wrong
There are two obvious moves, and both are mistakes.
The first is to take the work at face value. It’s excellent, so you treat the person as excellent. You hand them more, faster, with less oversight. You start mentioning them in rooms they’re not in yet. And then the scaffolding falls away — in a live client meeting, in a hard conversation, in the moment where there’s no tool between them and the problem — and the person is exposed. You promoted capability that wasn’t fully there. That’s not their failure. It’s yours.
The second is to quietly discount the work. It’s probably mostly AI. So you mentally dock it, withhold the credit, keep them in the holding pattern. But this patronises someone who has genuinely done the work of orchestrating a sophisticated piece of thinking. Knowing what to ask for, recognising when an answer is wrong, shaping rough material into something sharp — that is a real skill. It is not the same as producing the work unaided. But it is very far from nothing, and treating it as nothing is how you lose your best people to someone who sees them more clearly.
The honest position is uncomfortable
The honest position sits between those two, and it doesn’t resolve neatly.
The work is good. The underlying muscle isn’t fully built yet. Both are true at the same time.
Leadership now means holding both of those without flattening either. Not collapsing into “they’re brilliant, promote them,” and not collapsing into “it’s just the tool, ignore it.” Sitting in the genuine ambiguity and making decisions from there — about what to stretch, what to scaffold, where to let someone run and where to keep a hand on them.
That’s a harder cognitive task than the old one. It used to be a single judgment: is this good? Now it’s two judgments held in tension: is the work good, and is the person ready? They are no longer the same question, and a leader who keeps treating them as one will keep getting people wrong.
The same problem, from the other direction
It doesn’t only run downward. There’s a mirror version playing out at the senior end, and it’s just as unsettling.
Picture the person with fifteen years of pattern recognition. Whose value, for their entire career, has been that they could produce what others couldn’t. The deck, the analysis, the model, the document — that was the proof of seniority. Now there’s someone three rungs down who can match a good chunk of that output in a fraction of the time.
For that senior person, the ground has shifted. Seniority used to mean producing what others couldn’t. Increasingly it means judging what others produce. Knowing which of the four plausible answers is actually right. Spotting the flaw that looks like polish. Carrying the relationship and the risk that no model can carry.
Some will make that transition, and they’ll become more valuable than ever, because judgment at that level is rare and getting rarer. Some will quietly check out, defending an old definition of their worth that the work no longer supports. Part of leading now is helping experienced people make that move, before they decide the game has stopped being one they can win.
We’re teaching for a world that’s gone
Here’s what should worry anyone responsible for developing leaders.
None of this is in the programs. We’re still running feedback models designed in 2015. Still teaching people to assess output as though output still means what it meant. Still building capability frameworks on the assumption that the work and the person move together.
The actual job of leading has changed underneath us, and most of our development hasn’t noticed. The new job rewards a different muscle: the ability to read a person separately from their output, to hold two truths without forcing a resolution, to coach experience into judgment rather than production. Almost no one is being trained for that, because almost no one has named it yet.
I don’t think anyone has this fully worked out. I’m not sure anyone does. The leaders I work with are encountering it in real time, one ambiguous document at a time, mostly without a vocabulary for what they’re feeling.
But I’m fairly certain of one thing. The leaders who figure this out — who learn to separate the work from the person, and to lead both honestly — will be the ones worth working for.
The rest will be fine.
Until they aren’t.
Catalyst Enablement Group helps leaders and L&D teams build the judgment the new landscape actually rewards — separating capability from output, and developing the human advantage that AI doesn’t replace.
