The Most Expensive Search-and-Replace in Corporate History

Commentary3 min readPublished 2026-02-26AI Primer

Source: Jack Dorsey on X

AI and EconomyLabour MarketMarket Narratives
Cover image for The Most Expensive Search-and-Replace in Corporate History

Jack Dorsey announced today that Block is cutting nearly half its workforce — from over 10,000 to just under 6,000. The letter to employees is, by layoff-letter standards, remarkably well done. The severance is generous. The communication channels stay open so people can say goodbye. He writes:

We're not making this decision because we're in trouble. Our business is strong.

And then:

We're already seeing that the intelligence tools we're creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company.

So the story isn't "we're in trouble." The story is "AI changed what a company is." That's a much bolder claim, and Dorsey is making it with the confidence of a man who has clearly decided that this is the narrative and he's not hedging.

Here's what I keep coming back to: every big tech CEO has access to the same models. The same APIs. The same tooling. If AI productivity gains genuinely justified eliminating 4,000 roles at Block right now, the implication is that every company of similar size and shape is dragging around the same dead weight and just hasn't admitted it yet. Maybe that's true. But if it were obviously true, this wouldn't be news — it'd be an industry-wide pattern. It's not. It's one guy making a very large bet and calling it a conclusion.

I had two options: cut gradually over months or years as this shift plays out, or be honest about where we are and act on it now.

This is the move I'd flag. Two options. Not three, not five — two. "Slow bleed or rip the bandage off." It's a clean rhetorical frame, and it conveniently eliminates every moderate path from the conversation. Targeted restructuring, retraining, redeployment, phased transition — none of these exist in the binary. That's not honesty. That's stagecraft.

I don't doubt that AI changes how companies operate. I don't doubt that Block was overstaffed — most companies that hired aggressively in 2021–2022 were. What I doubt is the causal arrow. "AI made these roles unnecessary" is a very different statement from "we over-hired during zero interest rates and AI gives us a forward-looking justification for the correction." The first makes you a visionary. The second makes you every other CEO in tech circa 2023, just two years late.

The severance — 20 weeks plus tenure, accelerated vesting, six months of healthcare, devices, and a $5,000 transition payment — is real and worth acknowledging. The decision to keep Slack open for goodbyes instead of doing the silent-lockout that has become industry standard is a small thing that matters. "I'd rather it feel awkward and human than efficient and cold" is the best line in the letter, and it's the one that has nothing to do with AI.

The part that should make everyone pay attention is the product vision buried near the end: customers "building their own features directly, composed of our capabilities and served through our interfaces." That's not a staffing decision. That's a platform thesis. And if it works, it implies that this round of cuts isn't the last — it's the first. The logical endpoint of "customers build their own features" is a company that needs dramatically fewer people to build features for them.

Dorsey is not wrong about the direction. He's probably not even wrong about the destination. The question is whether you frame the largest layoff in your company's history as the courageous first step of a visionary — or whether you admit that the future arrived while you were still staffed for the past, and you're fixing that now, and it's painful, and AI is part of the reason but not the whole reason.

He chose the first framing. It's a better story. That's what makes it worth reading carefully.

Stay current weekly

Get new commentary and weekly AI updates in the AI Primer Briefing.