1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

Session 14 of 16 — AI Explorer

Explorer · Session 14 of 16⏱ 25 min · 👨‍👩‍👧 Parent present

🎨 AI and Images

Describe a scene in words — and see what AI creates.

🎯 Today's goal: Your child discovers how AI generates images from text descriptions — and learns that precision of language determines quality of output.

🔥 Warm Up — 2 minutes

Warm Up

Ask: "If you had to describe your bedroom to someone who had never seen it — so precisely that they could draw it exactly — what would you say?"

"Today we learn how AI generates images from words. The more precise your description, the better the result."

🤖 The Activity — 15 minutes

Main Activity

For this session, use an image generation tool — Canva AI (free tier) or Microsoft Designer (free) work well on a parent account.

🔧 Set up: Open Canva.com on your parent account. Go to AI Image Generator. Or use designer.microsoft.com — free with any Microsoft account.
📋 Type this exactly
A golden retriever puppy sitting in a field of sunflowers at sunset, looking directly at the camera, soft warm light, realistic photography style

See the result. Ask your child: What did it get right? What is different from what you imagined?

📋 Type this exactly
Now your child writes a description of something they want to see — their favourite animal in an unusual place, a scene from their favourite story, an imaginary place. They write the description. You type it in.

See the result. Improve the description together — what words would make it more accurate? Try 2–3 iterations.

The key lesson: every word in your description is an instruction. More specific = more accurate. This is the same principle as all prompt engineering — just now you can see it visually.

🧩 The Twist — 5 minutes

Find the Flaw
📋 Type this exactly
A cat sitting on a chair

Deliberately vague. See what comes out. Then refine: add colour, breed, style, lighting, mood, setting — one element at a time. Compare each version. Ask: What single word change made the biggest difference?

🛡️ Safety Moment
“Never share personal information”

Image generation from text is powerful and creative. The same rule applies: no real names, real locations, or identifying details go into the prompts. "A girl with red hair in a garden" is fine. "My daughter [name] at [address]" is never fine.

👨‍👩‍👧 Parent: read this aloud. Ask your child to repeat it back.
🎨

Image Maker Badge 🎨

Used text descriptions to generate images — and learned how precision changes the result

✓ Ask your parent to mark this session complete
👨‍👩‍👧

Parent Notes — tap to expand

▼
What they learned

Visual prompt engineering — understanding the relationship between textual precision and visual output. The iterative refinement process builds patience and careful observation.

Questions to ask
  • What single word made the biggest difference to your image?
  • What was hardest to describe accurately in words?
  • If you were designing an AI image tool for children — what rules would you build in?
What to watch for

This is a high-engagement session that children often want to extend. Keep a light eye on content — image generators occasionally produce unexpected results. If anything looks odd, just move to a different prompt.

Safety in context

Image generation tools occasionally produce unexpected or unusual content. Have a clear plan: if an image looks wrong, close it and try a different prompt. Do not dwell on it or show alarm — just redirect.

← Session 13Session 15 →