From Hierarchy to Intelligence: How AI is Reshaping Organizational Design

For two thousand years, companies have organized around a fundamental constraint: humans can only manage three to eight people effectively. This “span of control” limitation forced organizations into hierarchical structures, from Roman legions to modern corporations. Block is pioneering a radical alternative—replacing human coordination layers with AI-driven intelligence systems.

The Hierarchy Problem

Traditional organizational design stems from military necessity. The Roman Army created nested hierarchies (8 soldiers → 80 → 480 → 5,000) because commanders needed to route information across vast distances with limited communication. Prussian military reformers later added middle management—staff officers whose job was processing information and coordinating across units.

This military model entered business through American railroads in the 1840s. West Point-trained engineers brought hierarchical thinking to private companies, creating the first organizational charts to manage complex operations safely.

The constraint remains unchanged: more layers mean slower information flow, but growing organizations need coordination mechanisms. Every organizational innovation—from matrix structures to Spotify’s squads—has attempted to work around this tradeoff without breaking it.

The Intelligence Alternative

Block is questioning the core assumption that organizations must use humans as coordination mechanisms. Instead of giving everyone AI copilots to make existing structures work better, Block is building the company as an intelligence system.

This approach requires two components:

Company World Model: A continuously updated understanding of all operations, built from the digital artifacts that remote-first work creates. Every decision, discussion, code commit, and design exists as recorded data. AI maintains the complete picture that managers traditionally carried between layers.

Customer World Model: Rich transaction data from both sides of millions of financial interactions daily. When people spend, save, send, or borrow through Cash App and Square, they reveal honest signals about their financial reality. This data compounds into predictive models of customer and merchant behavior.

The New Architecture

Block builds four core elements instead of traditional product teams:

Capabilities: Atomic financial primitives like payments, lending, and card issuance. These building blocks have no user interfaces—they provide reliable, compliant functionality that other systems compose.

World Model: The dual understanding of company operations and customer reality, replacing information that flowed through management hierarchies.

Intelligence Layer: The system that composes capabilities into solutions for specific customers at specific moments. When a restaurant’s cash flow tightens before a seasonal dip, the intelligence layer automatically composes a short-term loan and surfaces it proactively.

Interfaces: Delivery surfaces like Square, Cash App, and bitkey where composed solutions reach customers.

When the intelligence layer cannot compose a needed solution, that failure signal becomes the roadmap—customer reality generates the backlog directly.

Three Roles Replace Hierarchy

This architecture requires only three types of roles:

Individual Contributors: Deep specialists who build and operate specific system layers. The world model provides context that managers traditionally supplied, enabling autonomous decision-making.

Directly Responsible Individuals (DRIs): People who own cross-cutting problems for defined periods, with authority to pull resources from any team. A DRI might own merchant churn in a specific segment for 90 days, then move to solve different problems.

Player-Coaches: People who combine building with developing others. They replace traditional managers whose primary job was information routing, focusing instead on craft and people development.

No permanent middle management layer exists. The system handles coordination, alignment, and priority-setting that hierarchies traditionally managed.

Why This Matters Now

Companies move fast or slow based on information flow. Hierarchy impedes that flow by design. For the first time, AI can perform the coordination functions that required human management layers.

The key question every company faces: What does your organization understand that is genuinely hard to understand, and is that understanding deepening daily?

If the answer is nothing, AI becomes just cost optimization. If the answer reveals deep, compounding knowledge, AI doesn’t augment your company—it reveals what your company actually is.

Block’s answer is the economic graph: real-time observation of financial behavior across millions of merchants and consumers. This understanding compounds every second the system operates.

The Transformation Ahead

Block is early in this transition, and parts will break before they work. But the pattern—organizing companies as intelligence rather than hierarchy—will reshape how organizations operate.

The Roman limitation of human span of control no longer constrains us. Eight soldiers still need coordination, but humans aren’t the only option for providing it. Block is building what comes next.