Monthly · September 2026 · S2⏱ 25 min · 👨👩👧 Parent present
✍️ Describing to Create
Write descriptions so precise that AI makes exactly what you imagined.
🎯 Today's goal: Your child masters visual description — the skill of translating imagination into words precise enough to direct AI.
🔥 Warm Up
Warm Up
Pick a scene: a rainy street at night, a forest at dawn, a busy market, a calm beach at sunset.
Today you write the description. AI draws. You evaluate.
🤖 The Activity
Main Activity
📋 Type this exactly
I am going to give you a visual description and I want you to: (1) identify any ambiguous parts, (2) ask me the questions you would need answered to make the image precise, (3) then write the most detailed possible image generation prompt based on my answers.
📋 Type this exactly
My scene is: [child describes their chosen scene]. Ask your questions.
Child answers AI's questions. Then:
📋 Type this exactly
Now write the complete image generation prompt using all the detail I gave you. Include: style, lighting, colour palette, perspective, mood, time of day, and any specific elements.
Generate the image. How close is it to what the child imagined? Refine with 2-3 more attempts.
🧩 The Twist
Go Deeper
📋 Type this exactly
What is the one type of image that AI consistently generates badly — something it almost always gets wrong? Why?
Common answer: human hands and faces at unusual angles. Ask: why is this harder for AI than generating a landscape?
🛡️ Safety Moment
“AI makes mistakes — always check”
AI image generation produces plausible-looking but sometimes factually wrong images. A generated "photo" of a historical event may look real but be completely invented. Never use AI images as evidence of something real.
✍️
Visual Storyteller Badge ✍️
Mastered the skill of visual description — precision in language producing precision in images
👨👩👧
Parent Notes
▼
What they learned
Visual communication and precision in language.
Questions to ask
What question AI asked helped you most?
What word made the biggest single difference to the image?
Can you think of something that would be impossible to describe precisely enough?
What to watch for
Precision in language is one of the most transferable skills from this session.
Safety in context
AI image safety: output can occasionally be unexpected. Close and try a different prompt if anything looks odd.