Understanding AI

Warm-up: Explain AI in two minutes

2 warm-up questions

• Has anyone ever asked you to explain AI — a colleague, a parent, a manager? How did it go?
• First try: I'm your manager and I've never used ChatGPT. Explain to me what AI is and how a tool like ChatGPT works — 30 seconds, go.

Reading: What your listener is really judging

Read a 190-word text about What your listener is really judging

What your listener is really judging When a colleague asks "so what actually is AI?", they are not asking for a computer-science lecture. They're silently judging three things: clarity (can you say it without jargon?), honesty (do you know what it can't do?) and relevance (what does it mean for me?). The classic failure is the jargon dump — "it's a neural network trained on a large corpus" — which sounds impressive and explains nothing. The opposite failure is hype: "it's going to change everything." Neither earns trust. Here is all the background you need to talk about it well. AI is software doing tasks that normally need human intelligence — understanding language, recognising images, making predictions. Most modern AI is machine learning: instead of following written rules, the system finds patterns in millions of examples. A tool like ChatGPT is a large language model — it learned the patterns of language from a vast amount of text, so it can predict what words come next. That's three sentences. Everything else is examples, analogies and honest limits — which is exactly what the structure on the next slides gives you.

Vocabulary: AI Vocabulary

6 items

Framework: The Plain-English Explanation

Define – Compare – Example – Limit, 4 steps

Define – Compare – Example – Limit: A four-move structure for explaining any piece of technology to a non-technical listener in under two minutes. The order matters: definition before analogy, example before limits — so your listener always has something familiar to hold on to.

Reading: Three ways to say it out loud

Read a 319-word text about Three ways to say it out loud

Three ways to say it out loud The same four moves — define, compare, example, limit — for three different listeners. Notice how the analogy and the register change, but the structure doesn't. To your manager (deciding whether the team should care) "In simple terms, AI is software that finds patterns in huge amounts of data and uses them to make predictions or generate text. Think of it like a very fast junior analyst who has read everything but checks nothing. You've seen it already — the spam filter on our email is AI, and so is the tool the sales team uses to draft follow-ups. The catch is accuracy: it's confident even when it's wrong, so for us it's a drafting tool, not a decision-maker." To a worried parent or friend (no tech background at all) "Basically, it's a computer program that learned from millions of examples instead of being given instructions. It's a bit like how you learned to recognise faces — nobody gave you rules, you just saw thousands of them. You use it every day without noticing — when your phone unlocks by seeing your face, that's AI. What it can't do is think or want anything. It's a tool, not a mind — the real risks are about how people use it, not robots taking over." To a client asking about an AI feature in your product (sceptical) "Put simply, the feature reads each incoming request and predicts which category it belongs to, the way an experienced employee would sort their inbox. Think of it as a first-pass assistant — it does the sorting, your team does the deciding. A good example: it can flag the urgent complaints within seconds instead of hours. To be honest about the limits — it's right most of the time, not all of the time, which is why every flagged case still gets a human review. That's deliberate."

Practice: AI Concepts

Gap-fill exercise, 5 items

• ____ learning means systems improve from data and experience.
• An ____ is a set of rules for solving a problem.
• ____ networks are inspired by the human brain.
• ChatGPT is an example of an ____.
• ____ an AI means teaching it using data.

Practice: Match AI Terms

Matching exercise, 5 items

• Machine Learning
• Neural Network
• LLM
• Algorithm
• Training

Dialogue: Explaining AI

Role-play: A colleague asks about AI basics (6 lines)

• Colleague: Everyone's talking about AI. What exactly is it?
• You: AI is technology that lets computers do things that usually need human intelligence - like understanding language, recognizing images, or making decisions.
• Colleague: How does it learn? Someone programs everything?
• You: Not exactly. Machine learning is when the AI learns from data instead of being explicitly programmed. You show it millions of examples and it finds the patterns.
• Colleague: What about ChatGPT? How does that work?
• You: That's a large language model, or LLM. It was trained on huge amounts of text, so it learned how language works and can generate human-like responses. It's basically very sophisticated pattern matching.

Discussion: Your two-minute explanation

3 discussion questions

• Mission: explain an AI tool you actually use (or your company uses) to me — I'm your non-technical manager. Use all four moves: define, compare, example, limit. Success = one analogy, one example I recognise, one honest limitation, and zero untranslated jargon. (Tutor: play the manager, time it — two minutes maximum, and interrupt with 'sorry, what does that mean?' if any jargon slips through.)
• Second take: I've just read a scary headline and I ask, 'So should we be worried about this stuff?' Answer in under a minute — balanced, honest, no hype in either direction.
• Compare this with your 30-second version from the start of the lesson. Which jargon did you drop, and which analogy are you keeping for real life?