ClinicianCraft
← Back to the course

FREE LESSON · MODULE 01

What generative AI actually does

Your goal: Distinguish a fluent draft from a verified answer.

Generative AI creates new text, images, audio, or other content from patterns learned during training and information supplied during use. A language model can produce a useful draft without having checked whether every statement is true. Its confident tone is a writing style, not a measure of reliability.

Think of an AI assistant as a drafting partner. You define the task, supply appropriate material, inspect the output, and decide whether it can be used. A clinician remains responsible for professional decisions. A receptionist remains responsible for confirming booking information against the actual diary. Neither responsibility transfers to the model.

For this course, use only fictional examples and material you are permitted to share. You can complete the exercises in any text-based AI assistant available to you, or compare your own written response with the model answer. There is no requirement to buy an AI subscription.

A useful first experiment is to ask for a short outline, then change only the audience. Notice how vocabulary and emphasis change. That flexibility is useful for communication, but it can also introduce unsupported details. Always compare the final output with the original task and source material.

TRY THIS PROMPT · FICTIONAL DATA ONLY

Create two outlines for a fictional clinic welcome guide: one for new staff and one for adult patients. Use only these facts: the clinic opens Monday to Friday, 9 AM to 5 PM; appointments require advance booking. Do not invent services or contact details.

Your turn

List three differences between the outlines. Circle any fact the assistant added that was not supplied.

Compare with the model answer +

The staff version may organize duties; the patient version may explain booking. Both must preserve the opening hours. Claims about emergency services, insurance acceptance, or same-day availability would be unsupported additions.

This is one complete lesson from the course. All exercises use fictional material.

Explore the full course ↗