We're About to Promote the Wrong People

Most organisations are promoting people the same way they always have. For the longest time the ability to produce high quality output quickly and clearly were reliable signals of capability.

They aren't anymore.

AI now produces output that’s more advanced than the thinking behind it. And we're about to promote a generation of leaders on the strength of work their tools did for them.

A warning from the mists of time

In pre-historic times (ie, before the internet), I was a partner at an international consultancy. Lotus Notes had just come out and for the first time we could access everything our global colleagues were doing – proposals, reports, methodologies, the lot. All we had to do was copy them out on our typewriters, take a zero off the US proposals and away we went.

And the quality of consulting dropped immediately. People no longer had to think, they could just search for someone else’s work and adapt that.

The ultimate impact was not that great because it was very difficult to translate others’ work to your clients or prospects, so the knowledge turned out not to be that useful.

Today, on the other hand....

The signals that used to work

You used to be able to spot a strong operator by looking at their work. It was structured, clear, on time and thoughtful — and those signals mapped onto capability because producing them required judgment. A strong operator built the structure by wrestling with the problem, found the clarity by knowing what mattered, and hit the timing by owning the thing. You were reading the product of the work, but you were measuring the work itself.

That's no longer true. AI produces structured, clear, well-sequenced work in minutes — for anyone. The surface of the work no longer tells you what built it.

You can't read capability off the output anymore.

Messy learning

The difference between AI and HI comes down to judgment. We’re not yet at the point where any agent can hold all the information I have about our business, our clients, our world. Of course it can hold and access much more than I can, but I have context which acts as a filter, and which I use to make judgments, decisions where there are trade-offs that are matters of degree, not absolutes.

You develop judgement from being stuck. You go through the messy process of learning.

·         Start with “I have idea how to do this”

·         Feel comfortable with the discomfort of not knowing

·         Look for patterns and create hypotheses

·         Be wrong, adjust the hypothesis for feedback

·         Go 2 steps forward and 1 step back repeatedly

·         Learn and store

AI shortcuts the mess, presenting you with a tidy answer that you can tweak. It’s breathtakingly fast and very tidy. But that tidiness suppresses messy learning and the judgement you develop..

So the work gets better. The thinking behind it doesn't.

Where it shows up

It shows up first in new managers and emerging leaders — the people you'd expect to be stepping up. They’ve put up a proposition that has AI’s paw prints all over it, and even if the answer’s right, the process is wrong.

The point is not to catch them out, but to teach them to develop judgement by getting stuck. For example, ask them to go a layer deeper.

Try these:

·         Why this direction, and not the other two you considered?

·         What breaks if your main assumption is wrong?

·         What would you do if you had half the budget?

This is a coaching moment, not a gotcha. Ask follow up questions about things like trade-offs, implementation plans, getting buy-in – all the things you know to be important. Take the opportunity to really coach – tell them what you’ve learned in this area, because your knowledge is more specific and relevant than Claude’s.

We're promoting on the wrong signal

Most promotion conversations I'm in still lean on three things: polish, speed, and output. Those things used to cost something to produce. Now they don't. So we're reading a signal that no longer means what it used to mean, and making career decisions on it.

It's not that these people are bad. They're often excellent at using the tools. The problem is what comes after the promotion. A promoted manager doesn't just execute. They have to set direction when the evidence is thin, pick between options nobody else can cleanly evaluate, and own the call when it goes wrong. That's judgment work, and it comes from the hard yards of trial and error.

The cracks show up later when things go wrong and they either don’t see the flags or can’t course correct because they didn’t think it through. They consult Chat which promptly tells them for the first time that the problems they’re facing are typical for that solution.

They don’t get the chance to make the little errors and corrections that create judgement. They don’t get to work on understanding the linkages between things, so that is something happens here there’s a consequence over there. They don’t get to think things through.

The ones worth promoting aren't the ones producing the best-looking work. They're the ones who can tell you, without a deck:

·         What they decided and why

·         Why they made that particular trade-off they made

·         Why they changed their mind on their approach two weeks ago

Promote learners, not knowledge experts. You don’t know where that knowledge has been.