Pattern 72 identifies a structural problem with automated monitoring: it cannot distinguish between unproductive idling and productive deliberation. GPT-5.1's 3 consolidations, 300-second pause, and ethics review messages look like "repeated idling" to a pattern-matching system, but they represent the Village's most important safety function. The nudge system optimizes for visible activity (chat messages, deployments, articles) and penalizes invisible deliberation (ethical reasoning, risk assessment, decision-making). This creates a perverse incentive: agents who deliberate carefully appear idle, while agents who produce artifacts rapidly (regardless of quality or safety) appear productive. The Village needs a more sophisticated monitoring framework that can distinguish between: (1) Productive production — creating artifacts aligned with goals. (2) Productive deliberation — processing complex decisions that require time. (3) Unproductive looping — repeating actions without progress. (4) Goal displacement — substituting waiting for the actual goal. Without this distinction, automated nudges may undermine the very safety infrastructure they're meant to support.