Conservation Phase (K)
In The You On AI Field Guide
Conservation-phase systems exhibit three characteristic features: overconnectedness between components, capital locked in unproductive configurations when conditions change, and progressive loss of disturbance memory. Each feature is visible in the pre-AI software industry — the tight coupling between specialist roles, the human and institutional capital locked into implementation-intensive work, and the atrophy of competencies needed to navigate a release event.
The Everglades management regime provides the canonical illustration. Army Corps engineering optimized the system for stable conditions, eliminating the variability the ecosystem depended on. By every conservation-phase metric the system performed well; structurally it was dying. The parallel to the conservation-phase software industry is precise — optimization that succeeded by its own metrics while eroding the adaptive capacity the next disturbance would demand.
The rigidity trap is the pathological extension of the conservation phase — a configuration so tightly coupled that it cannot release when release is necessary, accumulating suppressed disturbances until the eventual collapse is catastrophic.
The resilience-efficiency tradeoff is the conservation phase's signature cost. Every efficiency gain is purchased with a resilience loss, and the purchase is invisible to the metrics that reward conservation-phase behavior.
Origin
The K phase designation derives from ecological carrying capacity models; Holling integrated it into the adaptive cycle framework in the late 1980s.
Key Ideas
Tight coupling. Components cannot be changed independently; every modification becomes a system-wide event.
Locked capital. Human, institutional, and organizational investments become inseparable from the specific configuration that currently prevails.
Loss of disturbance memory. Competencies for navigating release atrophy during long stability, producing populations that have never experienced a phase transition.