Direction Coordination

In the AI Story

The distinction between production and direction coordination maps onto ancient philosophical categories. Techne — the knowledge of how to make things — is increasingly AI-performable. Phronesis — practical wisdom about what to make and why — remains human. Aristotle distinguished the two in the Nicomachean Ethics, observing that techne aims at an external product while phronesis governs action itself. The AI age makes this philosophical distinction economically operational: firms that organized around techne (production capability) are being repriced; firms that organize around phronesis (judgment about what to produce) retain their rationale.

Peter Drucker identified the knowledge worker's fundamental challenge as self-management: defining one's own contribution, setting one's own standards, evaluating one's own effectiveness. The challenge intensifies in the AI age because the knowledge worker's contribution can no longer be measured by output volume — AI produces abundant output. Contribution must be measured by the quality of the questions asked, the problems identified, the directions chosen. This is direction work, and it requires what Drucker called "the discipline of effectiveness": asking not "am I doing things right?" but "am I doing the right things?" The AI-augmented organization is one in which every member must exercise this discipline, because the tool will efficiently execute whatever direction is given.

The vector pod organizational form documented in The Orange Pill is the structural embodiment of direction coordination separated from production. A small group talks to users, analyzes markets, debates strategy, produces specifications that AI-augmented individuals or AI systems execute. The pod's value is measured by decision quality: whether the things it chose to build were the right things. This is expensive coordination — it requires senior judgment, cross-functional perspective, the tolerance for ambiguity that strategic decisions involve — but it is the only coordination that the AI age makes indispensable. The rest has migrated to software, to individuals, or to markets.

Origin

The explicit distinction between production and direction coordination as analytically separate functions appears to have crystallized in 2025–2026 practitioner discourse as organizations confronted AI's selective automation of managerial work. Michael Christen's analysis of the unbundled manager formalized the distinction, but the underlying insight is Coasian: the firm exists to reduce coordination costs, and when AI absorbs production coordination, only direction coordination justifies the firm's continued existence. The concept has roots in Henry Mintzberg's work on managerial roles and in the strategy literature's distinction between operational effectiveness (doing existing work well) and strategic positioning (choosing which work to do).

Key Ideas

Judgment cannot be delegated to systems without stakes. Strategic direction requires commitment — the willingness to be wrong and accept consequences — that computational optimization does not involve.

The inversion of organizational ratios. Traditional firms were majority-production, minority-direction; AI-augmented organizations invert to majority-direction, minority-production as execution becomes cheap and choice becomes precious.

Direction as the new scarcity. When anyone can build, the bottleneck migrates from execution capacity to the ability to decide what deserves building — a form of scarcity migration with direct wage implications.

The social infrastructure of good direction. High-quality strategic judgment requires trust, institutional memory, cross-functional dialogue, and the time to deliberate — precisely the social functions markets cannot provide and that justify the firm's persistence.