Agentation vs Agentification

Save an hour. Or rebuild the company.

January, 2026

Every Friday for the last two years, I spent an hour stitching together a status update. Slack threads, Linear tickets, half-written notes in my inbox. Last month I handed it to an agent. It now drafts the update, I edit for ten minutes, and I have my Friday afternoons back.

That is agentation. You take a task you already do, hand it to an agent, and reclaim the time. The shape of your work is unchanged. You are still the one doing the job, just faster. This is what most companies mean when they say they are using AI: a copilot for everyone, the same org chart underneath, slightly better margins on the same activities. Useful. Real. Finite.

Agentification is a different verb. It does not ask which tasks an agent can do. It asks what the hierarchy is actually for, and whether the system can do that instead.

What the hierarchy is for

In a recent essay, Jack Dorsey and Roelof Botha trace the org chart back to the Roman Army. Eight soldiers under a decanus. Eighty under a centurion. Five thousand under a legate. The Prussians added the General Staff in 1806 to support generals who could not hold the whole picture in their heads. The American railroads imported the structure in the 1850s. McKinsey formalised the matrix in 1959. Spotify, Zappos, Valve all tried to escape it and all reverted, or stalled, at scale.

The reason is the same one the Romans hit. A leader can effectively manage between three and eight people. Past that, you add a layer. More layers mean slower information flow. Two thousand years of organisational design has been an attempt to work around this trade-off without breaking it.

Their point is that hierarchy is not really about authority. It is an information routing protocol. A manager exists to know what is happening across their team and relay that context up and down the chain. The org chart is the wiring diagram for that relay.

What changes when the wiring is no longer human

Block is betting that AI is the first technology powerful enough to replace that protocol. In a remote-first company where decisions, code, designs and progress already exist as machine-readable artifacts, an AI system can build and continuously update a model of the entire business. What is being built. What is blocked. Where resources are allocated. What is working. The information the hierarchy used to carry, the system carries instead.

When that happens, the org chart inverts. The intelligence used to live in the people, and the hierarchy routed it. Now the intelligence lives in the system, and the people are on the edge. The edge is where the model meets reality, where intuition, ethical judgement and the feel of a room actually matter. Block collapses to three roles: individual contributors who build the system, directly responsible individuals who own cross-cutting outcomes, and player-coaches who develop people while still doing the work. No permanent middle management, because the thing middle management was doing has been absorbed by the model.

This is agentification. Not agents doing tasks inside the existing structure, but agents replacing the reason the structure existed.

The honest test

Agentification is currently more thesis than result. Block is in the early stages and openly expects parts of it to break before they work. It requires a foundation most companies do not have: machine-readable operations and a customer signal honest enough to compound. Block has both, the economic graph of millions of merchants and consumers, observed in real time. Most companies do not.

Dorsey and Botha end with the question every company should be asking. What does your company understand that is genuinely hard to understand, and is that understanding getting deeper every day? If the answer is nothing, AI is just a cost optimisation story. You trim headcount, lift margins for a few quarters, and eventually get absorbed by something smarter. If the answer is something, AI does not augment your company. It reveals what your company actually is.

Agentation is the cost-optimisation story. You can start this afternoon. Agentification is the reveal. It is a multi-year bet on a new organisational form, and it only pays off for companies whose understanding compounds.

Most teams will only ever need the first verb. A few are quietly working on the second.