For grown-ups

What's in it

8 lessons · about 44 minutes of reading · 33 check questions · 12 applied scenarios

1. 1. What Is Actually Happening — An AI chatbot is a prediction machine, not a know-it-all.
2. 2. When It Makes Things Up — Wrong answers arrive looking exactly like right ones.
3. 3. What Never Goes In The Box — The chat window is a postcard, not a diary.
4. 4. Help Versus Handing It In — AI can be your tutor. It cannot be the author.
5. 5. It Learned From Us — A computer answer can be unfair and still sound objective.
6. 6. Fakes, Voices and Scams — Stop inspecting the pixels. Check the channel.
7. 7. It Sounds Like It Cares — Agreeable is a product feature, not a friendship.
8. 8. You're The Human In Charge — Once you press send, it's yours.

Using it in a classroom

• One lesson per session works well: read the cards together, take the check as a class vote, then argue about the matching Judgment Calls.
• The scenarios are deliberately arguable. The disagreement is the lesson — ask students to justify a “risky” choice before you reveal why it's risky.
• Lesson 4 (schoolwork) says your rule beats this site's rule. Tell them your rule first, then let them compare.
• House Rules prints cleanly to one page — useful as a homework task to complete with a parent.

Conversation starters that actually work

• “Show me something an AI got wrong this week. How did you spot it?”
• “What's the rule in each of your classes about using AI? Which teacher is strictest?”
• “If a video of me asking you for money turned up, what would you do?”
• “What's something AI is genuinely brilliant for, and something people use it for that you think is daft?”
• “Has a chatbot ever agreed with you when you were actually wrong?”

Privacy, plainly

• No accounts, no sign-in, no email address, no cookies set by us.
• No analytics or advertising trackers, and no third-party scripts beyond the two web fonts served with the page.
• Lesson progress, quiz scores and House Rules are stored in the browser's own local storage on that device. Clearing site data or pressing “Reset progress” erases them.
• There is no chatbot embedded here, so children never type anything into an AI while using this site.

Where the site takes a position

• AI as a tutor is encouraged; AI as a ghostwriter is not. The test offered is “did the thinking happen in your head?”
• Chatbots are presented as useful tools that are tuned to be agreeable — not companions, and never a substitute for a trusted adult when something is wrong.
• Spotting fakes by eye is treated as unreliable. Children are taught to verify through a trusted channel instead.
• Any child showing signs of distress is pointed at a real person: parent, teacher or school counsellor.