🎉 Plan Something Fun
Use AI to plan your perfect birthday party.
🔥 Warm Up — 2 minutes
Ask: "If you could have the perfect birthday party — what would it be like? You have 2 minutes. Tell me everything." Listen and take notes.
"Let's see what AI adds to your ideas."
🤖 The Activity — 15 minutes
Your child is the client. AI is the event planner. Your child's ideas come first — AI adds to them.
Read together. Mark: ✅ keep this, ❌ no, 🤔 maybe. Your child decides.
Your child keeps this plan. If their birthday is coming up — show it to the family. AI helped, but the decisions were theirs.
🧩 The Twist — 5 minutes
Give AI an impossible brief:
AI will try to plan it anyway. Read the result together. Ask: "Is any of this actually possible? What would you actually do if this was real?" This teaches that AI always tries to answer — even when the question is unrealistic.
In a session like this — fun, creative, personal — it can be easy to forget the rules. Reinforce: even when AI is helping with something fun, if it says anything that feels strange or uncomfortable, we close the tab immediately and tell a parent. No explanation needed. No shame. Just close it.
Planner Badge 🎉
Used AI to plan something real — and made it better with your own ideas
Parent Notes — tap to expand
▼AI as a practical brainstorming tool — helping think of options they wouldn't have thought of alone, while the child makes all the real decisions. Also: giving AI context (brief) produces more useful results.
- What did AI suggest that you genuinely hadn't thought of?
- What did you decide NOT to use from AI's suggestions?
- If you actually used this plan, what would you change from what AI said?
This session tends to be a favourite — children get genuinely engaged. Watch for any tendency to share real personal details (real home address, real school friends' names) in the brief. Redirect: use made-up names for friends, don't specify actual locations.
The 'close and tell' rule is important to revisit in a fun session — because children are most likely to ignore safety instincts when they are enjoying themselves. Casually remind: 'If AI ever says something weird, we just close it, yeah?'