AI Layoffs Are a Leadership Failure

I was sitting in my hammock chair this morning when I came across an article in India Today about the falling morale at Meta. It's about what Zuck has been doing with AI — what it means for the people who work there, and what it says about the direction a lot of these companies are heading.

If you haven't been following this story, it's been going on for a while. Meta has been tracking what their employees type. Their mouse movements. Everything they do at a keyboard. And training their AI on all of it — not as a side effect, not as a pilot, but as the explicit plan to replace thousands of people. Eight thousand, by the current count.

What makes this different from the usual layoff announcement isn't the scale. It's how explicit they're being about it. Most companies reach for a cover story — economic headwinds, restructuring, a market that's shifted. Meta is saying out loud: we are training our AI on our people so that we can replace our people.

Atlassian went a different direction but arrived at the same place. Their version: lay off a large group of people, capture the salary savings, and use that money to invest in AI. And this wasn't even a case of "we've seen productivity improve dramatically, so we can do more with fewer people." There's no demonstrated outcome. No measurable gain to point to. It was just a bet: cut first, build the justification later.

I'm currently writing a book called What a Leader Owes Their Team. It's grounded in contractual ethics — the idea that when you step into a leadership role, you're not just adopting a management philosophy. You're taking on actual obligations. A debt you owe to the people who show up for you. And I've been deep in this thinking, because what I keep watching is a fundamental failure of exactly that.

I've been in leadership for over 20 years. And this pattern — outsourcing the hardest decisions to technology — is not new. Managers have been trying to do it forever. Since the first spreadsheet, there have been quantitative analyses designed not just to measure productivity but to create cover for decisions that are fundamentally human. Because here's how that logic works: if you make a call and it's the wrong call, you have to own it. But if the spreadsheet said so? If the algorithm recommended it? If the committee approved it? Then nobody's really at fault when it goes sideways. The accountability gets diffused until it disappears.

Except thousands of families are still living with the consequences. Real people struggling to pay their grocery bills and their mortgage. Kids who now have food insecurity — who won't be able to stay in the activities they were in, who might not get as good an education, who might have to move and lose their community and their friends. None of that gets absorbed into the model. The consequences don't disappear because a spreadsheet made the call.

And here's the part I find genuinely hard to stomach: the leaders making these decisions aren't even winning on their own terms. The companies going this route — their shares are actually down. They're worth less than they were before the cuts. So not only have they failed their teams, they haven't even delivered for their shareholders. They're chasing the bottom line, looking for shareholder investment, getting their bag — and they're failing all the way around. And then, usually about a year or two later, a lot of those same leaders get pushed out themselves. By then, thousands of other people have already paid the price.

I want to be honest: there's a part of me that wants to watch it blow up for them. The schadenfreude is real.

But then I think about those kids. And it stops being satisfying.

There is something almost indefinable about what a person brings to their work — their thinking, their imagination, their ethics and values and the soul that drives what they do. You can train AI on every action they've ever taken and still not capture any of it. Because those actions are an outflow of who they are. The AI learns what they did. It doesn't get access to why. It doesn't capture their discernment, the part of them that senses something is wrong before the data makes it obvious. Meta built what it built on top of those people. You can't replace them with an echo of their keystrokes.

I use AI heavily. In my own business, it's increased my productivity by about 8x so far, and I think that number will keep growing. But it hasn't replaced a single thing I do that actually matters. My ability to identify and define the right problems. To know what not to build. To read what a person or a team actually needs in a given moment, and respond to that. AI makes my work faster. It doesn't do the work that counts. And you can absolutely use AI to help your people do better, faster work — that is genuinely good. What you cannot do is replace the people with it.

At the end of the day, what do we actually have but people? What else matters — beyond our families, our kids, the people we spend our lives with, and our own souls? Does an extra zero in the bank account change any of that?

I know, I know — none of those leaders are probably reading this. But I want all of us thinking about it. About how we lead within our own context, our own companies, our own teams. About what it means to take on real obligations when you take on a leadership role. About what it looks like to actually become the kind of leader your people deserve.

That's what I write about here. The act of becoming. Choosing who you want to be — and then doing the hard work that proves it.

I'm glad you're here.

– Matthew