The AI Dividend Picks a Direction
Cloudflare just put AI on the receipt for 1,100 layoffs at record revenue. That makes the productivity story real and the 'AI creates more jobs' argument harder to sustain. The dividend has a direction.
Eleven hundred jobs at Cloudflare. Matthew Prince said it on the record, in plain English, attributed to AI efficiency in customer support. Not a euphemism. Not a restructuring narrative. Not a transformation story.
Six months ago, the honest take on most "AI layoffs" was that they weren't really about AI. The cycle was AI washing: a cost decision dressed up in a productivity story. That call still holds for plenty of companies. It doesn't hold for this one. And the New York Times reporting on Meta's internal AI mandate, which landed the same day, is the next domino in the same row.
The honest cut
TechCrunch covered the announcement straight. Cloudflare hit record revenue and laid off support staff in the same breath, with the CEO putting AI on the receipt. This is the inverse of the AI-washing pattern. Productivity gains in tier-one support are real, they're measurable, and at least one CEO is willing to say so on a public call.
I read this as a cleaner data point than anything we've gotten in the last year. Tech leaders have been whispering some version of "we found double-digit headcount reduction after the rollout" for months. Cloudflare just printed it. That makes it harder for every other CEO to keep the same observation in private.
Read the stock, not the headline
But here's where the framing splits. Yahoo Finance covered the same day differently. Cloudflare stock sank, and not because investors disagreed with the AI story. It sank because the revenue forecast missed expectations. Record current revenue, soft forward guidance, layoffs attributed to AI, all in one announcement.
Strip the framings: a guidance miss got bundled with an AI productivity story, and the productivity story did some of the work a guidance miss usually does alone. Both things can be true. AI did make those roles redundant. AI is also a useful narrative cover when growth decelerates in the same quarter.
A CEO attributing 20% of the workforce reduction to AI gains has an obvious incentive to do so. It tells the AI story shareholders want to hear and reframes a deceleration into a margin-expansion thesis. Take the productivity claim seriously. Discount the framing for the microphone.
What Meta is actually telling you
The NYT piece on Meta's 78,000 employees is the same story from the inside. Top-down mandates to use AI tools. Preparation for layoffs of those who don't adapt fast enough. Cultural pressure to attribute productivity gains to AI even when the AI didn't carry the weight. The reporters lean on the misery angle, and the misery is real. But the piece worth reading is structural.
A company is conditioning its workforce to ratify the productivity story before the layoffs come. That's how the dividend gets locked in. The narrative work happens before the headcount work. By the time the announcement lands, the AI attribution is pre-built and culturally enforced.
The "creates more jobs" argument needs a discount
The conventional wisdom on technology and labor is that historically, automation creates more jobs than it destroys. It's the line every reassuring op-ed reaches for and every economist on a podcast repeats. On a long enough timescale it's been true.
Here's where I'd push back. The argument doesn't say anything about who the jobs land with, when they land, or where the productivity gains go in the meantime. Cloudflare's gain shows up as margin. Margin shows up as buybacks and earnings beats. The eleven hundred people don't show up in the historical aggregate for several years, if they show up at all in the same income decile they left.
Call this the dividend's direction. The productivity is real. The direction is up, not across. That's a different argument from the "AI destroys jobs" panic, and a different argument from the "AI creates jobs" reassurance. Both miss the question that actually matters: who collects.
What I'd bet
By the end of 2026, "AI made the role obsolete" will be a standard line in tech-sector layoff announcements at profitable companies, and the SEC will start asking about whether the attribution holds up. The first company forced to walk back an AI-attribution claim under disclosure pressure will be more interesting than Cloudflare.
The dividend already picked a direction. The question is how long the framing holds before someone has to defend it under oath.
Sources
Want to talk about this?
Get in touchMore on AI
The AI Equity Story Sorts Itself By Q1 2027
Michael Burry's largest disclosed position right now is roughly $912 million in Palantir puts. The reasoning isn't that AI is fake. The reasoning is that the market hasn't sorted what's real from what's narrative.
The Ratepayer Subsidy Powering AI
Maryland just asked FERC to stop a $2B grid bill from landing on residential ratepayers to fund out-of-state AI data centers. It's the AI infrastructure story enterprise buyers should be tracking and aren't.
The Speed Mismatch on AI
CNN says Wall Street is finally pressing on AI returns. The NYT op-ed page says federal policy isn't moving fast enough on AI risk. Two pieces, two sections of the paper, opposite framings, same underlying story. The systems around AI weren't built to operate at AI's clock speed.