In 2002, a baseball team (the Oakland Athletics) with almost no money beat teams with ten times the budget by noticing that everyone else was measuring the wrong thing.

The story became a Michael Lewis book, and then a movie with Brad Pitt. Most people remember the story as being about math and statistics. I remember the temptation the team resisted. The cheap, obvious move was to cut the expensive veterans and move on. The harder choice was to look at players nobody wanted and see value that was sitting in plain sight, mispriced.

That latter choice is the whole story in a nutshell. Seeing value where everyone else has stopped looking.

Most companies that are having AI layoffs right now are not making that move. They are making the first one. They are cutting the expensive veterans, calling it innovation, and hoping the spreadsheet rewards them for it.

It usually doesn’t.

Over the past few months, I watched good people in my network lose their jobs. Strong operators who hit their numbers and brought value to their companies. For the companies brave enough to actually say it publicly, the reason was almost always the same two letters. AI.

The numbers back up what I was seeing. Through May of this year, US employers blamed AI for 87,714 job cuts, more than all of last year combined. Challenger, Gray & Christmas, which has tracked layoffs for decades, now calls AI the single most common reason companies give for cutting jobs.

Let me be clear about one thing: As much as I love using it and am fascinated by it, I’m not here to pretend that the technology is harmless. Jobs will be lost over the next few years. That part is real. What bugs me is the approach. I have been sitting in on events and peer-group rooms over the past few weeks, and this keeps coming up. The speed of the change. And the quieter thing underneath it, the moral weight of letting people go because someone decided a machine could do the job. (Normally, someone who has no clue how the job is done.)

To put it bluntly, layoffs are the lazy answer. They feel like decisive leadership. Most of the time, they are a failure of imagination.

Why then do companies turn to this playbook? Don’t they have some sort of obligation to their employees and the community? Seemingly, it is only to their investors.

More importantly, what about you? What do you owe yourself in this situation?

Deep stuff, I know. Let’s jump in.

Why AI Layoffs Cost More Than They Save

A layoff is the cleanest-looking line on a spreadsheet. One number goes down, and all of a sudden, it looks as if you have a brilliant business. The trouble is that the number rarely stays down.

Jeffrey Pfeffer at Stanford has studied this for years. He describes the current wave as social contagion, companies imitating each other. One competitor announces, the rest follow, and “AI efficiency” becomes the phrase everyone uses to sound modern and get that investor love.

Sandra Sucher at Harvard spent two decades measuring what the cut actually returns. Surprisingly (or not), layoffs rarely deliver the savings leaders expect, because the quick cut gets eaten up by lost knowledge, weaker engagement, lower innovation, and the exit of the people you wanted to keep. One study she cites found that trimming headcount by 1 percent led to a 31 percent jump in voluntary turnover the next year.

Basically, you let a few people go, and your best performers quietly start interviewing.

That essentially equates to a company selling its own future at a discount to make this quarter look clean. We live in a mad world at the moment, but this is also a quite dangerous form of madness.

When the cut is wrong, everyone finds out fast now

Klarna, the payments company, shrank its workforce by nearly half, mostly through a hiring freeze, while its AI assistant absorbed what the company said was the workload of 700 agents. For a while, it was the face of the lean, automated future. Then the CEO had to publicly reverse that decision. “Cost unfortunately seems to have been a too predominant evaluation factor”, Sebastian Siemiatkowski admitted, “what you end up having is lower quality.” His new line: “Really investing in the quality of the human support is the way of the future for us.” Klarna started hiring people back.

The Commonwealth Bank of Australia moved even faster from cut to apology. In July of last year, it declared 45 service roles redundant, pointing to a voice bot it said had reduced calls by 2,000 a week. The union checked. Call volumes were actually rising, and staff were being asked to work overtime, and managers were picking up the phones. Within a month, the bank reversed the redundancies, admitted an “error,” and apologized.

Now, I’m not saying every AI efficiency cut will be reversed. There are plenty that stick because there are areas of work that would most certainly be better done by a machine. But the pattern is loud enough that Forrester now predicts 55 percent of employers who cut jobs because of AI will regret it, and that roughly half of those layoffs will be undone in some form. A strategy that you need to go back on half the time is not a strategy. Now I’m no Jim Collins, Roger Martin, or John Kotter, but I’m sure they would all agree that that mental model is probably not the best business strategy.

Employee Experience Drives Profit, Not Layoffs

The truth is, whether we like it or not, treating employees well pays off. It is one of the better-documented profit engines we have. At my previous role, we were very interested in how the employee experience drove business value. Research shows that 86 percent of consumers will leave a brand they trusted after only two poor experiences, and that customer satisfaction depends almost entirely on how well a company equips its people.

If we look at it from the other side, Gallup found that disengaged employees cost their company the equivalent of 18 percent of their salary, while the most engaged teams are 14 percent more productive. It has become an age-old story. But the principle is simple and clear. Cared-for employees create cared-for customers, and cared-for customers create profit.

Even with that equation in mind, smart leaders keep reaching for the cut anyway. Leading a large company is genuinely hard, and the pull is real. Your decisions answer more to investors than to the people on your payroll, and investors want a clean quarter now. I have sympathy for that pressure. I’ve seen so many in these positions go grey way too early. Unfortunately, it’s also exactly where the laziness is hiding. In many cases, cutting jobs satisfies the loudest voice in the room at the time, but it rarely satisfies the business 18 months down the road.

The young pay the highest price, because they live on wages

The question I get more than any other right now is about people just starting out. What happens to the 22-year-old when the entry-level positions are the first thing AI is told to climb?

Scott Galloway framed the underlying unfairness a few years ago. People who earn through wages, mostly the young, get squeezed, while people who hold assets, mostly the already comfortable, do fine. AI layoffs aimed at junior roles pour fuel on exactly that fire. You remove the rung where judgment gets built, then wonder why, in five years, nobody on the team has any. There are ways around this, but even many of those just push the buck further up the ladder.

And what we’re talking about isn’t a “small” ask or change. The World Economic Forum estimates that 39 percent of today’s core skills will be outdated or transformed by 2030, with most workers needing real reskilling to keep up. That includes your skills and mine. A churning skills problem on that scale is an argument for equipping people.

You see, a layoff also never ends at the office door. If you’re reading this, you may come from a small enough town where one big employer can quietly hold up a whole neighborhood. The bakery on the corner. The youth club. The tax base that pays for the library. Cut deep enough, and all of it feels the cut. There is a line I saved years ago from a Harvard Business Review piece that talks about this: “your computer doesn’t unemploy you, your robot doesn’t unemploy you. The companies that have those technologies make the social policies that change the workforce.” AI is not going to fire anyone. A real person makes that decision.

The work that’s safe is the work that actually uses you

At one of those events, a young lady had an interesting presentation. She had joined a big consulting firm, expecting her analytical skills to be the thing that mattered. Her strengths pointed somewhere else entirely. Connection. Adaptability. Energy. She realized that she had maybe been (harshly) grading herself on the wrong test.

The reframe she needed is the one most of us need now. Success is not whether you can do the work. It is whether the work lets you use what you are actually good at, the part an LLM can’t hand back to you.

That same HBR analysis found the hardest work to automate is managing and developing people, and applying hard-won judgment to messy decisions. Taste. Trust. Reading a room. Owning the call when it goes wrong.

IKEA understood this with the same robot everyone else was buying. Its bot, Billie, ended up handling close to half of all call-center queries. That is the exact moment most companies reach for the lever. IKEA reached the other way. It retrained roughly 8,500 call-center workers as remote interior-design advisers, a paid service it had never offered, and that channel generated 1.3 billion euros in a single year. Same people. Newer, more valuable work. Even IBM, hardly a soft touch on automation, told the Wall Street Journal that after AI took over a chunk of HR tasks, “our total employment has actually gone up,” because the savings funded new roles rather than eliminating old ones.

The technology was identical. The imagination was not.

A few ways to prepare (or How to AI Proof Your Job)

You can’t control which kind of leader signs your paycheck. You can control how ready you are before the dreaded conversation reaches your desk.

I think about this quote from Winston Churchill a lot.

“To each there comes in their lifetime a special moment when they are figuratively tapped on the shoulder and offered the chance to do a very special thing, unique to them and fitted to their talents. What a tragedy if that moment finds them unprepared or unqualified for that which could have been their finest hour.”

I’ve always had that in the back of my mind should an opportunity arise. At this point in time, AI has brought a vast amount of opportunity, but with it will certainly come some unexpected taps on the shoulder. You need to be prepared for those situations as well. Preparation will decide whether those moments find you ready.

Here are a few ways I think you could prepare.

Know what’s durably yours. Spend an afternoon looking at where your week actually goes. Some of it an AI model can already do. Some of it the AI only helps with. And some of it stays yours. Things like the judgment calls, the relationships, the read on which answer is actually right. That last part is your edge against the machine, and as we just saw, it is also the hardest thing to automate. Most people have never looked at their own work this way, and you can’t protect what you can’t see. Like that young consultant, the trick is to grade yourself on the right test.

Point your learning at the edge, not the tool. Get genuinely fluent in AI, but aim that fluency at your strengths instead of chasing every new feature. The goal should never be to keep up with the machine. It’s to become the person who runs it. A sharper, better version of you, doing the work that only you can do, because now the grunt work is done by the machine.

Tend your relationships before you need them. I have been laid off twice. Both times, what carried me wasn’t my title or my tools. It was the people who picked up the phone. Relationships are the most durable human asset you have, and the least automatable. Build them when you don’t need anything, so they are there when you do. Time and life may get in the way, but be kind and don’t be an asshole. People will always help you out.

And if you lead people, do the same: take an honest look at your team before you ever approve a cut (if you have a choice). If you can’t name the part of someone’s job that’s durably theirs, you haven’t looked hard enough yet. It is the same honest look behind the AI Skills Audit I shared earlier, aimed at a team rather than a tool. That isn’t a people problem. It is a leadership one.

Brad Stulberg writes that change itself is neutral. “It becomes negative or positive based on how we view it and, more importantly, what we do with it.” AI is the change. The layoff is one thing a company can do with it. Reskilling is another. Equipping yourself is the one fully in your hands.

I think the young lady at that event will be fine. She’s aware now of what makes her awesome. Something that we should all aim to figure out. The people I watched get cut this year will mostly be fine, too, because the strong ones already know what is theirs to carry into the next thing.

That has probably always been the real test, long before AI gave it a new name. Know what is yours, and build toward it on purpose.

Layoffs are the lazy answer. The serious one is to make yourself impossible to answer with a layoff.