How To Teach AI: A Simple 4-Lesson Plan For Anyone
The short version
- Teach AI through the learner’s own tasks. A demo on someone else’s example rarely sticks.
- Four short lessons are enough: first chat, better asks, checking answers, and building a habit.
- Put checking and privacy rules in lesson one, before anyone pastes in a client file.
- Adjust the examples for the group. Your team, a client, and your parents need different tasks, and the same plan works for all of them.
- Measure success by one task each person now does faster.
If you’re figuring out how to teach AI to your team, a client, or your family, start with this rule: every learner uses their own real task from minute one. People learn AI by using it on something they care about, then fixing what it gets wrong. The lesson plan below runs in four sessions of about 45 minutes and works for groups of one to twelve.
Before You Teach: Three Prep Steps
- Pick one tool for everyone. Mixing tools in the first session creates confusion. Choose the one your group already has access to, such as a company-approved chat tool or a free tier they can sign up for.
- Ask each learner for two tasks in advance. Something they write often and something they read often. “Weekly status email” and “long vendor contracts” are perfect answers.
- Write down your ground rules. What data can go into the tool, and what can’t. Print it or share it on screen.
That second step does most of the teaching for you. When people see AI handle their own weekly email, you don’t need to sell them on it.
How To Teach AI In Four Lessons
Lesson 1: The first conversation (45 minutes)
Goal: everyone gets a useful result and knows the rules.
- 10 minutes: show one live example on a task from the group. Type slowly and talk through what you’re doing.
- 5 minutes: cover the ground rules. No client names, passwords, health details, or account numbers in the tool.
- 25 minutes: each person runs their own writing task. Walk the room and help.
- 5 minutes: each person shares one thing that surprised them.
Give them this starter prompt to adapt:
“I’m a [your role]. Write a first draft of [the task] for [the reader]. Keep it under [length]. Ask me up to three questions first if you need more information.”
The last line matters. It teaches people that AI can ask for context, which leads straight into lesson 2.
Lesson 2: Better asks (45 minutes)
Goal: learners see how context changes the result.
Have each person run the same task twice: once with a one-line prompt, once with the reader, the purpose, an example, and a length. Compare the two side by side. This single exercise teaches more than any slide about prompting. Our guide to writing AI prompts makes a good handout for this session.
End by showing follow-up edits: “shorter”, “warmer”, “cut the jargon”, “make it a table”. Many beginners think they get one shot. They don’t.
Lesson 3: Checking the output (45 minutes)
Goal: nobody trusts AI blindly.
Use the reading task this time. Have learners upload or paste a document and ask for a summary with dates and numbers. Then have them check three facts against the source. Somebody in the room will almost always find an error or a missing detail. That moment teaches the lesson better than a warning.
Share a short checklist, such as the one in how to spot AI mistakes, and ask everyone to keep it near their desk.
Lesson 4: Build the habit (45 minutes)
Goal: each person leaves with one saved workflow they’ll use weekly.
- 15 minutes: each person writes a reusable prompt for their most common task and saves it in a notes file.
- 20 minutes: they run it on a fresh example and refine it.
- 10 minutes: set a two-week check-in. Each person reports one task that now takes less time.
Adjusting The Plan For Different Groups
Your team at work
Get your company’s AI policy in writing before lesson 1. Use real but non-sensitive work, like meeting agendas, internal emails, or process docs. Pair a keen learner with a nervous one in lesson 1. After lesson 4, collect the best saved prompts into a shared document.
Clients
If you’re a coach or consultant, teach AI inside their business problem. A client who runs a bakery should practice on menu descriptions and supplier emails. Keep a written summary of what you covered and which prompts you gave them. It becomes a deliverable, and it protects you if a question comes up later about advice given.
Family
Shorter sessions work better, around 20 to 30 minutes. Use personal tasks: planning a trip, writing a thank-you note, understanding a letter from the insurance company. Spend extra time on scams. Teach them that AI can make fake voices and fake messages, and that a call asking for money needs a second check through another channel. For learners who want to keep going alone, AI tutor prompts give them a way to practice between sessions.
Mistakes Teachers Make
- Showing too many tools in one day.
- Spending the session on how AI works inside instead of what it does for the learner.
- Skipping privacy rules because “everyone knows”.
- Doing the typing for learners. Let them type, even when it’s slow.
- Ending without a follow-up date. Without one, most people stop using it within a week.
- Overselling. If you promise magic, the first mistake loses the room.
Knowing how to teach AI is a skill you can sell. Coaches, consultants, and office managers who can run these four lessons well become the go-to person for it.
Turn AI Skills Into Better Client Work
AI For Coaches And Consultants shows you how to use AI in your own practice and how to bring clients along: session prep, follow-up notes, client-facing materials, and a prompt pack you can hand over. It includes a worked example from intake to final report.
