If Your Child Could Build an AI, What Problem Would They Solve?

Ask your child to invent an AI, and you might hear about a robot that finishes homework or a machine that finds every missing pencil. Now ask a different question: who would it help, and what problem would it solve?

That second question is where this family activity begins. All you need is paper, a pencil and an everyday situation worth improving. No coding or AI account is required.

A recent idea worth bringing home

In an article published on 2 October 2026 and updated on 5 October, UNESCO described an education workshop that began with identifying problems before choosing technology. Participants also considered ethical implications before deployment. The article explains that defining the need can reveal where AI might help and where another approach would be more suitable.

The activity below is an original BAIO adaptation of those broad principles for families. It is not a UNESCO lesson plan, an official exam question or evidence that completing it improves test scores.

Start with one small, familiar problem

Invite your child to describe something that could work better at home or school. Make the problem specific enough to discuss: finding a misplaced water bottle, choosing a library book or understanding which bin an item belongs in.

Ask who experiences the difficulty and what a useful result would look like. A clear starting sentence might be: “I want to help students find their missing belongings without searching every classroom.”

Keep the first version small. A child can explore one situation in detail before imagining a system that solves everything.

Draw a possible solution in six boxes

1. Who needs help?

Draw the person who would use your idea. What are they trying to do? What makes the task difficult? For the missing-bottle example, the user might be a student who remembers the bottle's colour but cannot remember where they left it.

Discuss whose perspective you might have missed. A teacher managing lost property may have different needs from the child looking for an item.

2. What would the system do?

Describe one helpful action. Imagine a system that compares a description with pictures of items in a lost-property cupboard and suggests possible matches. This is a hypothetical design, not a claim about a working BAIO product.

Then ask whether AI is needed. A labelled shelf, a simple list or an adult helping students might solve the problem well. Choosing a simpler approach is a valid design decision.

3. What information would it need?

For the pretend bottle finder, useful details might include colour, shape and a drawing of a sticker. Use made-up items for this paper exercise. There is no need to collect classmates' names, photographs or personal information.

Ask what the system could do if the description is incomplete. Would it show several possibilities? Would it ask another question? The design should allow uncertainty instead of pretending every match is correct.

4. What could go wrong?

Two children might own similar blue bottles. A photograph might be unclear. A sticker might have peeled off. Discuss how a helpful suggestion could still be mistaken.

Draw a safeguard beside the idea: an adult checks ownership before an item is handed over. The system can suggest a match while a person checks the decision. This keeps the example focused on help, evidence and responsibility.

5. How would you test it?

Create four fictional bottle cards: two similar ones, one with a clear identifying feature and one that does not match the description. Ask another family member to play the system using the rules your child wrote.

Try an easy description first, then an ambiguous one. Record what happened and whether the rules need changing. This is a paper simulation of a decision process; it does not train or test a real AI model.

6. How would you know it helped?

Choose a practical sign of success. Perhaps the design narrows the search to two possible items, or clearly says when there is insufficient information. Also ask how you would notice a mistake.

Compare the proposed system with the simple lost-property list. If the list works better for this situation, explain why. A useful idea earns its place by meeting a need, rather than by having AI in its name.

Try a different idea together

If lost property does not interest your child, imagine a library helper that suggests books from a fictional reading list. Ask what preferences it should consider, how it could offer variety, and whether a librarian should help with the final choice.

Or design a pretend recycling guide using drawings of packaging. Ask who would verify the local sorting rules and what should happen when an item is unfamiliar. Keep these as discussion examples until a responsible adult has checked any real-world instructions.

The point is to explore a need, a possible approach and its limitations. There is no single winning invention.

Adapt the activity for Classes 3–8

For Classes 3–5, use pictures and short spoken explanations. An adult can write down the child's words. Focus on who needs help, what the idea does and one possible mistake.

For Classes 6–8, add detail about data, uncertainty and testing. Ask students to compare two approaches and explain what evidence would change their choice. They can also discuss who should review a decision and how someone could challenge a wrong result.

Allow about fifteen minutes for a first sketch, then continue only while the discussion is useful. This is a suggested family activity, not a scored readiness assessment.

Share the reasoning behind the drawing

Invite your child to present the idea to a family member: “This helps someone do this task. It needs this information. It could make this mistake, so here is how we would check.” Ask the listener to contribute one helpful question.

Continue with BAIO's five-minute family AI challenge or our parent guide to AI and learning, both linked below.

Your child's drawing does not need to become an app. Explaining the problem, considering alternatives and revising an idea already gives the family plenty to explore.