The Coordination Problem

In the AI Story

The coordination problem is so deeply embedded in organizational thinking that it is rarely examined as a specific problem with specific historical causes. Yet it is specific, and it is historical. Coordination becomes expensive when specialization is deep, when translation costs between specializations are high, and when the work requires integration across multiple specializations. These conditions characterized industrial production from roughly 1750 to approximately 2025 — an extraordinary run, but not an eternal one.

Orange management theory treated the coordination problem as permanent. Frederick Winslow Taylor, Alfred Sloan, Henri Fayol, and their successors built increasingly sophisticated apparatus for solving it. The result was extraordinarily productive and characteristically bureaucratic. Every organization built management hierarchy proportional to its coordination demands. Every career track led through management. Every business school taught the skills of coordination.

AI has not reduced the coordination problem. It has eliminated the premise on which the problem rested. When a backend engineer can build interfaces with AI assistance, when a designer can write features, when a product thinker can prototype directly, the specialization that generated coordination demand dissolves. The work that previously required a team of specialists can be done by a single person moving fluidly across domains. Coordination between specialists becomes obsolete when specialization itself becomes optional.

This dissolution does not eliminate all coordination needs. Organizations still coordinate across geography, across time, across customers, across strategic objectives. But the specific form of coordination that Orange hierarchies were built to provide — the coordination of specialized human execution — is no longer scarce. And the apparatus built to provide it, if retained after the problem has dissolved, becomes friction that drags on the organization's actual work.

Origin

The coordination problem as an explicit object of organizational thought dates to the early industrial revolution, when factory managers confronted the novel challenge of integrating specialized workers whose skills did not naturally mesh. Adam Smith's pin factory in The Wealth of Nations (1776) described the productivity gains from specialization; the coordination challenge those gains created became the central concern of industrial management over the following century.

Taylor's scientific management (1911), Fayol's administrative principles (1916), and Sloan's multidivisional structure at General Motors (1920s) represent the progressive sophistication of Orange solutions to the coordination problem. Each was a genuine advance; each assumed that coordination was a permanent organizational requirement. The AI revolution has revealed the assumption as historical rather than natural.

Key Ideas

Historical, not permanent. The coordination problem arises from specific conditions that can change.

Three preconditions. Deep specialization, high translation costs, integration requirements.

Hierarchy as solution. Orange management apparatus was built specifically to solve this problem.

AI dissolution mechanism. Cross-boundary execution by individuals eliminates the specialization premise.

Residual coordination needs. Purpose, strategy, geography remain; specialized-execution coordination does not.