The New Inefficiency

In the AI Story

The old soldiering, for all its limitations, contained a wisdom Taylor's framework could not recognize: the wisdom of limits. A human being is not a machine. Rest is not waste. The capacity for judgment — which the AI age has revealed as the only irreplaceable human contribution — degrades when exercised without interruption, just as a muscle degrades when loaded without rest. The informal work norms Taylor despised were collective agreements to limit production to sustainable levels. They were imposed not by individual willpower (which the tool's availability erodes) but by institutional structure (which persists regardless of individual temptation).

Segal's transatlantic-flight confession in The Orange Pill captures the phenomenology of the new inefficiency with confessional precision. The writer works through the night not because the deadline requires it but because the tool makes stopping feel like self-suppression. The exhilaration drains away; the grinding continues. The pattern of compulsion is visible to the writer, but the mechanism of interruption — the informal norm, the colleague at a neighboring desk, the flow state's natural termination — is absent. The tool is available, the work is possible, the internalized imperative converts availability into obligation. What Taylor called soldiering was, in part, a natural limiter. AI has removed it.

The Berkeley researchers documented the dynamic with empirical precision. Task seepage — the colonization of pauses, breaks, and marginal moments by AI-assisted work — is the organizational manifestation of soldiering's elimination. Workers did not lose their breaks because a manager took them away. They lost their breaks because the tool was available, the work was possible, and what Han calls auto-exploitation converted every available moment into a production opportunity. This is more insidious than anything Taylor designed, because it operates without external compulsion.

The correct measurement would track judgment quality over time rather than output per unit. Judgment, like every other human capacity, follows a curve — improving with engagement, peaking with focus, degrading with exhaustion. The builder who stops at a reasonable hour may produce less output but better judgment than the builder who works through the night. The output is higher on the dashboard; the judgment is invisible in the data; the dashboard wins. The organizations that recognize this will build what might be called institutional soldiering — the AI Practice framework the Berkeley researchers proposed — structured limits on production that protect human capacity for judgment against the unlimited demand the tool makes possible.

Origin

The framing emerges from the collision of Taylor's original framework with the empirical findings of AI-work studies beginning around 2023. The historical parallel — that the informal norms Taylor despised were replaced over the twentieth century by formal labor protections — is now being re-staged, as the protections themselves prove inadequate to the new technological conditions.

Key Ideas

The elimination mechanism. AI removes the natural and collective restrictions that embodied labor had imposed, making unlimited production operationally feasible for the first time in history.

Volume without value. Production past diminishing returns produces work whose quality degrades as the director's capacity degrades — the efficiency gain becomes efficiency loss as the curve flattens.

Auto-exploitation as mechanism. The new inefficiency operates without external compulsion; the worker enforces unlimited demand on herself, experiencing the enforcement as freedom.

The judgment curve. Human judgment follows a diminishing-returns curve with engagement duration; stopping earlier preserves the quality that extended work degrades.

Institutional soldiering. The contemporary equivalent of the eight-hour day — structured collective limits on production designed to protect judgment quality against the tool's continuous availability.

Debates & Critiques

Whether institutional soldiering can be built through voluntary organizational commitment or whether it will require the political struggle that produced the twentieth century's labor protections is the open question. Individual willpower has proved inadequate against the tool's continuous availability. Organizational commitment has proved fragile when quarterly metrics conflict with long-term cultivation. The historical pattern suggests collective action and perhaps legal intervention, on a timescale the compressed AI transition may not allow.