Adoption & Trust: Why Teams Either Avoid AI or Blindly Over-Rely On It
People do not cluster in the middle of AI competence — they cluster at the extremes: avoidance or uncritical reliance. That is the avoidance↔trust axis (five-stage continuum, transfer corpus).
3D: Vermeidung ↔ kritische Nutzung ↔ blindes Vertrauen
1. The extremes (Guide 1, §2.1)
- Avoidance: fear, technostress, lack of psychological safety.
- Uncritical reliance: LLM fallacy, over-attribution, "AI makes me competent".
Both are dysfunctional endpoints. The middle — critical, context-aware use — is the goal.
2. AI Mindset (triadic, empirically tested, C)
Disposition (attitudes, self-efficacy, mindset) × context × technology appraisal. Psychological factors measurably shape competence and use.
3. A-factor (N=517, C)
One dominant latent factor across interaction tasks (44% variance) — there is a coherent construct to train, not a loose collection of tool skills.
4. What this means for your program
- Objective measurement instead of self-report (see Self-report lies).
- Role-specific tasks instead of generic training (see Design for the Job).
- Train disposition: self-efficacy, critical evaluation, not just tool operation.
Further reading
Sources: Guide 1 (Implementation), §2.1/§2.2; corpus of 16,459 papers (as of 2026-10). EOF echo "en9 done"