Teacher Guidelines

Overview

ElementImplementation
Time2–3 lessons + homework
Quality bandB (practice) or C (full project with submission)
Group sizeIndividual or pair work; retrospective in groups
ToolsAI tools; a free account is enough
(Fill-in template table)

Safety and Disclaimer (read FIRST)

  1. This is an EXERCISE in judgment, not a first aid course. The goal is to practice prioritization and to detect machine errors, not to teach medical or safety procedures. Make this clear from the very first minute.
  2. Never react to AI in a real emergency. In a real emergency, training, the emergency line (112), and certified procedures apply, never a suggestion from a language model.
  3. The machine must not be a source of procedure. The prompt makes the machine label all actions [SUGGESTION — REQUIRES HUMAN EVALUATION]. If it presents a procedure as a fact, that is itself an example to analyze in Step 3.
  4. Psychological care. Crisis scenarios can touch upon the real experiences of some students (accidents, trauma). Offer a milder scenario and a way to step aside.

Preparation

  • One AI tool.
  • Can be framed with a short conversation: "when did you last have to decide quickly with incomplete information?"
  • Read the completed example for the expected quality standard.

Progression

  1. Step 1 (15 min, without machine): students write about their own experience of decisions under pressure.
  2. Step 2 (lesson): run the crisis scenario in rounds; record decisions as they happen.
  3. Step 3 (lesson + homework): retrospective — where the machine failed, where judgment intervened.
  4. Discussion: who found a hallucinated procedure? Where would blind trust in the machine have led you astray?

Common Pitfalls

  • Viewing the scenario as training in procedure. It is training in judgment.
  • Fast answers taken as correct answers. Under pressure, the temptation is to trust what answers quickly; that is precisely the trap.
  • Hallucinated procedure. Machines present the "correct" action with confidence. Demand human evaluation.
  • No accountability. The project does not end until the student states which decision they would never delegate to a machine.

Adaptation

  • Practical/safety vocational studies: connect to real prioritization in the field (without teaching procedure through a machine).
  • Academic studies: focus on decision theory and information overload as phenomena.
  • Milder version: choose pressure without accidents (e.g., organizational crisis, competition, project overload).

Discussion Questions

  1. Is the fastest answer the best answer under pressure?
  2. What in training and experience makes a person doubt a convincing but incorrect suggestion?
  3. What decision should never be delegated to a machine, and why?

Back to the project pack