The wildebeest thunder across the Serengeti. The caribou cross the Canadian tundra in their hundreds of thousands. And every July, with startling reliability, LinkedIn fills up with the annual Microsoft migration as hundreds of talented, experienced people leave at the end of the financial year.

I say this with genuine affection - I was part of it myself nine years ago. It’s not new and it’s not unique to Microsoft - every big tech company, probably every big company full stop, goes through some version of this cycle, every year, on schedule. This year it landed harder than usual, maybe because of the voluntary redundancies and the scale of them, but also because of how tough the market is for those newly arriving in it. But strip away this year’s specifics and you’re left with something that happens like clockwork: a wave of people with two or three decades of experience, moving on.

What catches my eye every time, more than the who or the why, is the what - the sheer amount of accumulated wisdom walking out the door in a single month. Not knowledge. Wisdom. The two get treated as the same thing, and they are not.

Knowledge is knowing the answer. Wisdom is knowing which question actually matters. It’s recognising a pattern three moves before it becomes obvious. It’s having enough scar tissue to smell a plausible-sounding bad decision before anyone has committed to it. It’s accumulated one difficult decision at a time, through success, failure, responsibility and consequence. And unlike knowledge, it can’t simply be downloaded or prompted into existence.

For most of my career, knowledge was the scarce resource. Organisations competed for people who knew things: people who understood the technology, the market, the process, the customer or the industry better than everyone else. We hired for that knowledge. We promoted for it. We paid for it. We built entire hierarchies around the people who possessed it.

Generative AI changes the economics of that world.

Facts, summaries, explanations, first drafts, analysis and competent-sounding answers to almost anything are becoming abundant and nearly free. Everyone with a connected device now has access to more instantly retrievable knowledge than many of the world’s largest organisations possessed a generation ago.

That does not make knowledge worthless. But it does make it less scarce. And economics has always taught us that scarcity shapes value.

As knowledge becomes abundant, the thing that becomes more valuable is not the ability to retrieve an answer. It is the ability to judge whether the answer is useful, relevant, ethical, commercially sensible or completely wrong.

The scarce resource in the AI economy is wisdom - the judgement to know what to do with everything knowledge now hands you for free. Which makes it particularly strange that so much of it appears to be migrating out of organisations at precisely the moment they should value it most.

You’d think that would make people like my former colleagues more employable, not less. It doesn’t seem to be working out that way. Some have found their next chapter quickly, but many others have faced months of applications, silence and rejection, despite carrying exactly the kind of wisdom I’ve just described. I don’t think that’s an accident. Most hiring systems were built in the knowledge era, to screen for knowledge: keywords on a CV, years in a specific tool, a title that matches a template. They were never designed to detect wisdom, because wisdom doesn’t keyword-match. So the CV of someone with thirty years of hard-won pattern recognition gets filtered out by a system built to look for something else entirely, before a human ever reads it.

That’s the real cost of the migration. It isn’t that wisdom is leaving - wisdom has always left, that’s what retirement, redundancy and reinvention have always looked like. The cost is that we’ve built organisations and hiring systems optimised for a resource, knowledge, that just became abundant, while the thing that’s now genuinely scarce walks straight past them, unrecognised.

None of this makes leaving easy. Moving on from something you have given decades of your life to can feel like a loss, particularly when the decision was not entirely yours. It takes time to find your feet again and to realise that the role, challenge or version of yourself that comes next may ultimately take you further than the one you have left behind. That has been true for me and for many others I know who have made the same journey. But it should not be made harder than it needs to be by organisations that have failed to update their understanding of value.

AI has changed the economics of work more profoundly than many leaders yet realise. When knowledge was scarce, hiring for knowledge made perfect sense. Now knowledge is abundant, the scarce resource is wisdom.

Organisations that continue to optimise their hiring, promotion and leadership systems for yesterday’s scarcity will become extraordinarily efficient at filtering out the very people they need most.

Wisdom was never a sentimental nice-to-have. It’s the thing that turns information into a good decision. Organisations that work that out first won’t just survive this year’s migration - they’ll be where the next wave of wisdom chooses to land.