Five ridiculous names for a laundry robot

Tonight’s tiny AI challenge does not need a project file or a quiet hour.

Ask your child to invent the prompt, not just request the answer.

Try: “Give us five ridiculous names for a robot that only folds laundry.”

Then let the humans argue over the winner.

Why a silly prompt is real practice

A lot of “AI help” is a child pasting a question and hoping. The skill that actually transfers is the opposite: deciding what to ask, how specific to be, and whether the reply is any good.

Common Sense Media’s family activities lean on this. Change the wording. Compare. Talk about why the answer moved. You do not need a lesson plan. You need a job small enough to finish before the kettle boils.

A laundry robot is perfect. Nobody’s marks depend on it. Everyone has an opinion about names. The fight at the end is the point — humans still choose.

How to run the ten minutes

1. They write the first prompt. No help except “make it more specific if you want.” Five ridiculous names is already better than name a robot. 2. Read the five out loud. Groan. Laugh. Cross out the dull ones. 3. Change one constraint. “Now they all have to include a sock.” “Now the robot is grumpy.” “Now the names must sound like a pirate.” Run it again. 4. Vote. One person, one vote, including you. Ties are settled by the person who last folded something.

If you want a second round, swap the job: a robot that only finds lost water bottles, or only packs the PE bag. Same shape. New argument.

What you are practising:

• Detail. “Ridiculous” and “only folds laundry” are instructions. The model follows them, or it does not.
• Taste. A list can be fluent and still boring. They get to say “that one is trying too hard.”
• Ownership. The prompt was theirs. The winner is a family decision, not the machine’s favourite.

What you say if they stall

Some children go blank when asked to “be creative.” Give them a stem, not the whole line.

• “Start with: Give us five…”
• “Add a rule the robot must obey.”
• “Add a rule the names must obey.”

If the answers are bland, do not blame the child. Blame the prompt, together. “We didn’t tell it to be messy enough.” That is a better lesson than “AI is amazing.”

Keep phones face-up. Keep it short. If it stops being fun, stop. The habit you want is I can steer this thing — and I do not have to accept the first list.

Want a harder round? Ask them to write a prompt that would make bad names on purpose — too long, too rude, too boring — and then fix it. Directing a tool includes knowing how you made it worse. That is the same muscle as “we didn’t tell it enough.”

If they want to draw the winner, let them. A robot named Sock Justice on a cereal-box body is a better souvenir than another screenshot.

Tomorrow they can use the same muscle on a real task: a tighter question for homework, a clearer ask for a picture, a prompt that does not leak the family address. Tonight is socks. Fold-o-matic 3000 can wait in the hallway with the rest of the laundry.