AI Productivity Ratchet / AI 生产率棘轮

AI productivity ratchet is the EP275 workplace pattern where people who use AI well may become busier because their improved output becomes a new baseline for task assignment. 陈明霞 describes the reversal of the “AI agent as top worker” story: the person who can coordinate AI may become the one who receives more work, more review responsibility, and more pressure.

The concept extends Ratchet Effect In The Workplace / 职场棘轮效应 into AI adoption. A worker may reasonably hesitate if AI productivity is rewarded mainly with higher targets, tighter deadlines, or more unpaid supervision. This makes Workplace Incentive Design and AI Job Security Anxiety part of adoption, not afterthoughts.

Key Claims

• AI productivity can become self-punishing when managers convert higher output into ordinary capacity without changing reward, scope, or staffing.
• The ratchet is strongest when AI work is visible as speed but hidden as review, correction, prompting, context-building, and responsibility.
• “Learn AI or be eliminated” narratives can hide incentive problems by treating reluctance as backwardness rather than risk assessment.
• AI rollout needs explicit agreements about productivity gains, role redesign, evaluation, and who owns mistakes.
• The pattern can contribute to AI Brain Fry when workers supervise more AI output than they can sustainably inspect.

Connections

• Ratchet Effect In The Workplace / 职场棘轮效应, Workplace Incentive Design, and Workplace Pacing - management and pacing base concepts.
• AI Job Security Anxiety, AI Worker Literacy, and AI Native Worker - worker-adoption and career-risk branch.
• AI Use Pacing, AI Brain Fry, and Human Judgment Under AI - review-capacity and responsibility branch.
• Human-Scale AI Use / 人作为 AI 的尺度 - evaluative boundary against productivity for its own sake.