The Timeline
On February 26, Jack Dorsey cut 4,000 people from Block and told Wall Street that intelligence tools have changed what it means to build and run a company. He said most companies were late and predicted that the majority would reach the same conclusion and make similar structural changes within a year.
He did not have to wait the year.
Twenty-four days. That is the gap between Dorsey making the cut and Zuckerberg building an AI version of himself to justify making his own.
The Playbook Is Identical
This is not two CEOs arriving at the same conclusion independently. This is one CEO watching the other be first, and be validated. Now he is scrambling to catch up.
Dorsey / Block
Zuckerberg / Meta
One column is the work of 18 months. The other is the work of three weeks. Dorsey built his tool, tested it in production, measured the results, then made the structural decision. Zuckerberg saw the stock pop, bought what he did not have, and is building his agent now.
Dorsey built it. Zuckerberg bought it.
The CEO Agent
The Wall Street Journal detail that matters most is not the agent itself. It is the description of what it does. The agent retrieves information Zuckerberg would normally have to go through layers of people to get.
Layers of people. That is what Dorsey just cut.
Zuckerberg is not automating his assembly line. He is automating himself. He is building a tool that replaces the human chain between a question and an answer at the highest level of a $1.5 trillion company. The CEO is the proof of concept. If the agent can skip layers for him, it can skip layers for everyone beneath him. And if it can do that, those layers stop being organizational structure and start being overhead.
Meta's CFO, Susan Li, said something at a conference earlier this month that tells you everything about the internal pressure driving this. She said she thinks a lot about making sure Meta does not work any less efficiently than companies that were AI native from day one.
An 80,000-person company is scared of 20-person startups. Zuckerberg is using AI to eliminate layers inside a trillion-dollar company. You are still routing decisions through three people to figure out if you need to order bacon. That fear is what produced a $2 billion acquisition, a personal CEO agent, and a reported plan to cut 16,000 people. All in a month.
The Market Is Voting
Every single company that has announced AI-driven workforce reductions in 2026 has seen its stock price go up. Block jumped 25%. Meta climbed on the layoff report. Atlassian gained after announcing cuts. Amazon. Salesforce. The pattern is not subtle.
Wall Street is not debating whether AI replaces jobs. Wall Street is rewarding companies that say it does and act on it. That is an incentive structure. Once the market starts paying a premium for AI-driven efficiency, every CEO with a board and a quarterly earnings call faces the same math. Cut and get rewarded, or wait and explain why you are behind.
Dorsey told his shareholders that he would rather act honestly and on his own terms than be forced into it reactively. Meta's response to that statement was to do exactly what he described: react.
The AI-Washing Debate Misses the Point
Half the media coverage of the last month has been about whether these layoffs are "real" AI replacements or just cost cutting in a tech-industry costume. A former Block communications head wrote in the New York Times that it was organizational bloat wearing an AI costume. An economist at the Federal Reserve wrote that Dorsey's prediction was wrong. Bloomberg ran a piece questioning whether it was genuine or AI-washing.
It does not matter.
Whether AI is the reason or the excuse, the structural outcome is identical. Smaller teams. Flatter orgs. Intelligence tools handling coordination, information retrieval, and reporting that humans used to do. If a CEO is cutting people and the stock goes up and AI gets the credit, then AI is the operating framework regardless of what the underlying motivation was. The tools are real. The restructuring is real. The fact that some of it might also be convenient narrative does not change what lands on the rest of us.
Where This Lands
The tools being built to solve Zuckerberg's problem do not stay inside Meta. Block's Goose is open-source. Meta's Manus is a subscription product for small and medium businesses. The internal tool a Meta employee built on Claude, described as an "AI chief of staff," indexes documents, queries project data, and talks to other agents on behalf of its user. Those capabilities do not stay in Silicon Valley. They end up in the software the rest of us run our businesses on.
When Square builds Square AI, when Toast ships ToastIQ reporting, when any POS platform adds an intelligence layer, they are building the same thing Zuckerberg just built for himself. Smaller. Simpler. But the same architecture: an agent that retrieves information you would normally have to go through layers of people to get.
For a restaurant, those layers are the shift lead checking the walk-in, the manager pulling a labor report, the bookkeeper reconciling invoices. The information already exists inside the systems. The friction is in getting it to the person who needs to act on it. That friction has a cost, and that cost just became visible to every CEO in the world because Dorsey measured it, cut it, and got rewarded for it.
The question stopped being whether AI replaces jobs the moment Zuckerberg started building an AI clone of himself to do his own job faster. The question now is how quickly the tools built for trillion-dollar companies show up at your front counter. And how the rumors of Mark being a robot grow.
Based on the last 24 days, the answer is: faster than anyone predicted. Including Dorsey.