On March 31, 2026, Jack Dorsey and Sequoia Capital's Roelof Botha published an essay called "From Hierarchy to Intelligence." Most companies saw the headline, noted the controversy, and moved on.

That was a mistake.

This wasn't opinion. This was a structural blueprint — written by a founder who had just eliminated 40% of his workforce and then explained, in precise detail, exactly why the corporate org chart is over. As someone who has spent 27 years building digital systems across 15 countries, who helped grow five companies to $1B+ in revenue or valuation, I can tell you: this is the most important thing published about business structure in 2026.

Here is what the essay actually says — and what every N5R.ai client needs to understand about it.

The Essay

Why Hierarchy Was Built — and Why AI Just Made It Obsolete

Dorsey traces the org chart back to the 1850s. Daniel McCallum designing the first corporate hierarchy for the Erie Railroad. Before that: the Prussian military. Before that: Roman legions.

Every hierarchical structure in history solved exactly one problem: humans can only hold and relay so much information at once. A manager exists because the CEO can't talk to 10,000 people. A director exists because the VP can't manage 200. Layer after layer, built entirely to move information up and down the organization.

Dorsey's argument is clean and devastating: AI just made that constraint irrelevant.

That is not a prediction about some distant AI future. That is a $2.87B gross-profit company — Block — already rebuilding around this premise after cutting nearly half its people.

The Architecture

What Block Is Actually Building

Most coverage fixated on the layoffs. The layoffs are the consequence. The essay is about what replaces the org chart.

Dorsey describes two AI systems he calls World Models:

The intelligence layer sits on top of both. It coordinates work, routes decisions, surfaces answers — the things a middle management layer previously existed to do. Three roles replace the traditional five-layer org chart:

• Individual Contributors — Deep specialists with direct access to the company model and the people who can act. No gatekeepers. No approval chains.
• Directly Responsible Individuals (DRIs) — Own cross-cutting problems on 90-day cycles. Full accountability, no committees. AI keeps them coordinated.
• Player-Coaches — Still build things. Also develop people. Management as a separate career track from execution is gone.

Block's current structure: five layers between Dorsey and any employee. Target: two to three layers within one year.

The Pattern

This Isn't Just Block — It's a Wave

Amazon cut 30,000 corporate roles in late 2025 and early 2026, explicitly framing it as flattening management layers. Meta's 2026 layoffs specifically targeted mid-level management. Shopify's CEO required AI competency as part of performance reviews — headcount dropped from 11,600 to 8,100 while revenue grew 20–40% annually. McKinsey says AI could automate activities accounting for 57% of US work hours.

This is showing up in earnings calls right now.

The Bigger Idea

Simulate Your Business Before You Run It

Here's where I want to go deeper than the headlines.

Forget org charts. Think about what a CEO actually does every single day. They run simulations in their head. Constantly. "What happens if I cut this price?" "If I approve this hire, what does Q3 look like?" "If we lose this client, can we hold the team?"

Every business decision is an educated guess. Now think about how Waymo solved the same problem for autonomous driving.

They didn't put cars on roads and hope. They built Simulation City — by 2026, running 10 million simulated miles per day across 25,000 virtual cars. In February 2026, they launched the Waymo World Model, built on Google DeepMind's Genie 3, capable of generating scenarios that don't yet exist in the real world.

The car doesn't guess. It has survived every possible situation before touching a real street.

We have never done this for business. We have dashboards, analytics, AI agents writing emails. But none of it answers the one question every CEO asks every morning: "If I do X, what actually happens?"

That is what Dorsey is building at Block. NVIDIA is building the infrastructure layer in manufacturing — Omniverse deployed at BMW, Toyota, TSMC. Siemens acquired Altair for $10B to go deeper on simulation. OpenAI took a stake in Thrive Holdings specifically to embed its teams inside traditional companies and demonstrate this at scale.

The company that can simulate the future will outcompete the company that reacts to it.

The N5R.ai Framework

What This Means for Your Business — 4 Moves to Make Now

I want to be precise about something. Block had $2.87B in Q4 2025 gross profit, proprietary transaction data from tens of millions of active users, and a remote-first culture already generating machine-readable artifacts from every decision. Their circumstances are specific. Your results will differ based on your infrastructure, market, team, and execution.

But the framework is universal. Here is what we are implementing right now at N5R.ai — and advising every client to start immediately:

Honest Counterpoint

What the Skeptics Are Right About

Block tripled headcount from 4,000 to 13,000 during the pandemic, then cut to 6,000. There is a credible argument that these cuts correct pandemic overhiring, with "AI efficiency" as a convenient frame. Sam Altman told CNBC directly: "There's some AI washing where people are blaming AI for layoffs that they would otherwise do."

Harvard Business School studied 50,000 developers and found AI reduced project management time by roughly 10% — real, but not the wholesale revolution the headlines imply.

And the best managers don't just route information. They read the room. They make the judgment call when data says one thing and instinct says another. That is hard to automate.

Hold both things at once. The structural argument is correct. The "AI replaced managers and everything is fine" narrative needs more evidence. The companies that win will be honest about that gap while building toward closing it. That is exactly what we help N5R.ai clients do.