Teaching · Practical guide

How to Use ChatGPT to Prepare a Lesson Plan

Start with what learners must demonstrate. Ask AI to suggest a sequence, then check the subject knowledge, timing and assessment yourself. This worked example uses Linux file permissions; adapt it to your own subject.

My DYAP →

Define the learning outcome before opening a tool

A useful outcome describes something observable: “Given a file permission string, identify who can read, write and execute the file.” “Understand permissions” is too vague to assess. Specify prior knowledge, class duration and available equipment.

For this example, assume adult beginners who can navigate a Linux directory and have access to a disposable practice environment. Use invented learner profiles, not names, grades or personal circumstances.

Use a bounded lesson-planning prompt

Copy the template below and replace the subject and learner details. Ask for a short first draft so you can inspect it before requesting slides or worksheets. Different AI tools may produce different results.

Worked example: a 60-minute lesson

Minutes 0–5: show -rw-r----- and ask learners to predict who can access the file. Minutes 5–15: explain owner, group and others, then the meaning of r, w and x for a regular file.

Minutes 15–25: model two permission strings aloud. Minutes 25–40: learners create a disposable file and compare permissions before and after a change. Minutes 40–50: pairs explain a deliberately incorrect answer. Minutes 50–60: complete an individual exit question and discuss corrections.

Example exit question: For a regular file with permissions -rw-r-----, can an ordinary user who is neither the owner nor in the file’s group read it, considering these mode bits alone? Answer: no; the “others” bits grant no access. State that this simplified example excludes ACLs and privileged access.

Review the draft before teaching

Check that the timings total 60 minutes and that the exit question tests the stated outcome. Verify commands in a disposable environment. Do not give students commands you have not checked.

Look for hidden assumptions: directory permissions behave differently from regular-file permissions, and a demonstration can depend on the current user and group. Add a simpler task for beginners and an extension requiring explanation rather than more commands.

Try it and record one improvement

Draft your own 30-minute lesson, then score it against four checks: accurate subject content, realistic timing, suitable difficulty and an observable assessment. Revise any failed check. Keep the original and revised versions to see where human judgement improved the draft.

Make the learning evidence visible at each stage

Before preparing slides, write the learner action beside each teaching action. “Explain permissions” describes what the teacher does; “identify the owner, group and others fields and justify an access decision” describes evidence you can inspect. If a planned segment contains only teacher talk, add a short learner response before moving on. Keep the response small enough to review while there is time to correct a misconception.

For the regular-file example, a useful sequence is to label the three permission groups, identify the relevant user category, then interpret that category’s bits. Do not begin by asking learners to memorise a command without explaining the decision it expresses. The lesson’s simplified permission questions explicitly exclude ACLs and privileged access; those limits should remain in the slide, task and answer key.

A learner who says “the file has read permission” has not yet specified for whom. A targeted follow-up is “Which three bits apply to the user in this scenario?” This keeps the feedback connected to the missing reasoning rather than treating every wrong answer as a need for more explanation of the whole topic.

Worked practice answer: adapt the plan to 30 minutes

One defensible 30-minute plan allocates 3 minutes to a diagnostic question, 7 to modelling, 10 to guided interpretation, 6 to an independent check and 4 to feedback. The times total 30 minutes. Remove the file-changing demonstration if there is insufficient time for a safe environment check; do not rush commands merely to preserve every activity from the 60-minute version.

Use the fictional permission string -rwxr-x--- for a regular file. The owner has read, write and execute bits; group members have read and execute bits; others have no permission through these mode bits. An ordinary non-owner group member cannot write the file on this simplified evidence. Ask learners to point to the group triplet r-x and explain the missing w, rather than merely selecting “no”.

For an independent variation, use -rw-r----- and an ordinary user who is neither the owner nor in the file’s group. Under the stated mode-bit assumptions, the others triplet --- grants no read access. This is not a claim that mode bits alone describe every real system’s access behaviour.

Self-review the adapted plan: do the times total 30 minutes; does every task test the specified outcome; and does the answer key preserve the scenario’s assumptions? Keep a fallback worksheet if the practice environment is unavailable. Record a change you would make after observing learners, without claiming that this illustrative plan has already been classroom-tested.

Reusable prompt

Select and copy this template, or customise your own using the free builder.

Act as a lesson-planning assistant. Plan a 60-minute introductory lesson on Linux regular-file permissions for adult beginners who can navigate directories. Outcome: interpret owner, group and other permissions accurately. Use a disposable practice environment. Return a timed table with teacher action, learner action and evidence of learning. Include one worked example, a common misconception and an exit question with an answer. State assumptions and items the teacher must verify. Do not include destructive commands.
Open the prompt builder →

Pause · Think · Check

Test your understanding

Choose one answer per question, then check your reasoning. This is practice; scores are not saved.

1. Which outcome is easiest to assess?
2. What should you do before teaching an AI-generated command?

Answer every question to check your results.

Sources and further reading

Continue learning

Browse this learning track →