The Competence Anxiety Loop

Here is the loop. The AI tool is deployed. The employee does not know how to use it well. They are aware that they do not know how to use it well. They are also aware that the organizational message is 'embrace AI' — which they interpret, correctly, as 'demonstrate AI competence.' Admitting they cannot use the tool feels like admitting incompetence in a context where incompetence is career risk. So they avoid the tool. The skill gap widens. The anxiety increases. The avoidance deepens.

This is not laziness. This is a rational behavioral response to an environment where vulnerability is punished.

1. Diagnosis Signal 1: Usage without engagement The employee logs into the AI platform regularly. Their work product shows no AI influence. They are running queries, getting results, and ignoring them. The gap between access and integration is the behavioral signature of the competence anxiety loop.
2. Diagnosis Signal 2: Task-specific avoidance The employee uses AI for low-stakes tasks — writing subject lines, summarizing documents they wrote themselves — and avoids it for the high-stakes work where it would produce the most value. The avoidance is specific to tasks where poor AI output would be visible. That is not random. That is threat-calibrated avoidance.
3. Diagnosis Signal 3: Competence comparison anxiety The employee monitors colleagues' AI use with unusual attention. Asks questions that are less about the tool and more about how well others are using it. This is competence benchmarking. They are trying to assess whether their gap is visible and how large it is relative to peers.
4. The Intervention Normalize the learning curve explicitly and publicly. The loop feeds on the gap between the cultural message ('embrace AI') and the learning reality ('AI takes time to use well'). Close the gap. Say out loud, in a venue with organizational authority: 'It takes 90 days to develop useful AI workflow habits. Everyone is at different points. This is expected.' Then make it true by tolerating imperfect early output without visible judgment.