Efficiency vs. Intensity

In the AI Story

The productivity metric cannot distinguish between these two sources of improvement. A worker who produces twice as much because the process became more efficient and a worker who produces twice as much because she is thinking twice as hard look identical in the statistics. The metric sees only the ratio. The mechanism is invisible. This blindness has always been present. It has rarely mattered as much as it matters now.

In an industrial setting, the intensity of labor was partially self-limiting. A factory worker operating at unsustainable intensity made errors, damaged equipment, suffered visible physical exhaustion. The limits were embodied. In a knowledge economy augmented by AI, the limits are cognitive and therefore invisible — invisible to the manager, invisible to the metric, often invisible to the worker herself until depletion manifests as burnout, reduced judgment quality, or the flat affect that the Berkeley researchers documented.

The distinction maps onto Segal's ascending friction thesis from the supply side. Where Segal argues that AI eliminates lower-order friction and exposes higher-order friction, Coyle's framework asks whether the higher-order engagement is sustainable or whether it represents cognitive expenditure being drawn faster than it can be replenished.

For the AI-revolution reader, the efficiency-intensity distinction is the single most consequential measurement gap in the current policy conversation. A metric that cannot distinguish them will reward the wrong pattern — celebrating firms that extract more from their workers while overlooking firms investing in sustainable transformation.

Origin

The distinction has antecedents in labor economics and industrial sociology but Coyle's articulation of it in the AI context appears in her Stanford Digital Economy Lab white paper on measuring AI (2024) and her work with Jörden and Poquiz on firm-level AI adoption (2025). The framework also draws on Byung-Chul Han's diagnosis of the burnout society as theoretical background.

Key Ideas

Observable identity. Efficiency and intensity gains look identical in aggregate productivity statistics — only cognitive load data distinguishes them.

Sustainability asymmetry. Efficiency gains compound indefinitely; intensity gains are borrowed from reserves that must eventually be repaid.

Embodied invisibility. Cognitive intensity limits are invisible to managers, metrics, and often to workers themselves until depletion manifests.

Policy distortion. A measurement system that cannot distinguish the two will systematically reward extractive patterns over sustainable ones.

Debates & Critiques

Some economists argue that current AI productivity gains are predominantly efficiency-based, citing examples where AI handles mechanical tasks and frees humans for higher-value work. Coyle's position is empirically cautious: the composition cannot be determined from current metrics. The Berkeley study's documentation of task seepage, reduced empathy, and flat affect among AI-augmented workers provides evidence consistent with intensity patterns, but longitudinal data over years rather than months will be required to settle the question.