A few months ago someone on our team finished in an afternoon a piece of analysis I would have budgeted two days for. I asked how she had done it. She walked me through it, and about halfway through I realized I was no longer reviewing her work. I was being taught. She was patient about it, and I was grateful, and I also felt a small, unfamiliar discomfort that I have been trying to understand since.
The discomfort comes from an old assumption about management: the person in charge can, at least in principle, do the job of the people reporting to them. I came up in finance, where that was close to literally true. You earned the right to review a model because you had built a hundred of them. That assumption never needed defending because it was almost always correct. Now there are whole categories of work where the newest person in the room has the most fluent hands, and the manager's experience is real but sits in a different place than the craft.
What the manager is still for
My first instinct was to catch up, to spend my evenings learning every tool so I could keep my footing. I did some of that and I do not regret it, but I came to see it as a partial answer at best. The tools change every few months, and a person whose job is also to hire, decide, and carry responsibility will not out-practice someone who uses them eight hours a day. Trying to win that race made me a worse manager for a while, because I started evaluating people by how they worked rather than by what they produced.
What I can still do, and what her speed does not replace, is tell her whether the answer is worth having. Her analysis was fast and clean, and it was also answering a slightly different question than the one the decision actually turned on. Nothing in the tool would have told her that. I knew it because I had been wrong in a similar way years ago and paid for it. That is the part of experience that survives: knowing which question matters, what a plausible-looking number is hiding, and what the cost is if the work is wrong. It is judgment about consequences, and it is not nothing.
Reviewing work you could not have made
The harder adjustment is reviewing output when you cannot reconstruct the process. I used to check work by retracing steps. Now I check by interrogating it: what did you assume, where would this break, how would you know if it were wrong, what did you verify yourself versus take as given. These questions work whether or not I understand the tool. And I have noticed that people who are truly skilled answer them easily, while people who are merely fast go quiet. That silence is a more reliable signal than anything I used to look for in a spreadsheet.
I have also made the mistake of being too impressed. Fluent output has a way of lowering a reviewer's guard, and for a stretch I approved things faster than I should have because they looked finished. Polish is now cheap, so it tells me very little. I have had to relearn that a confident, well-formatted answer deserves the same suspicion as a rough one, and perhaps slightly more, since it is less likely to invite scrutiny on its own.
Letting the teaching run both ways
What I have settled on is an arrangement that felt awkward at first and now feels ordinary. People who know the tools better than I do show me what they do, and I tell them what I am worried about in their work and why. Neither of us pretends the exchange is symmetrical. They are not my mentors on judgment, and I am not theirs on technique. But I think a manager who cannot say out loud, I do not know how you did that, show me, will end up with a team that hides its methods, and then no one is learning anything.
I am not certain this holds up as the tools get better. It is possible that the judgment I lean on will itself be partly automated, and that my usefulness narrows further than I would like. I would rather say that plainly than build a comfortable story around it. For now, the job looks less like supervising a craft and more like owning the question, the standard, and the consequences, while staying curious enough to be taught by the people closest to the work. That is a humbler version of the role than the one I inherited, and I suspect a more honest one.