Your Kid's First Podcast Episode

Children are asked to write constantly and to speak almost never. Yet speaking is where thinking gets tested — you cannot hide a half-formed idea behind neat handwriting.

This workshop produces a five-minute audio episode with your child as the host. It takes about an hour, needs no equipment beyond a phone, and uses AI for exactly one thing: helping them prepare good questions.

Last verified: August 2026.

Why audio, and why now

Speaking exposes understanding. A child who has written a paragraph about volcanoes may or may not understand volcanoes. A child who has talked about volcanoes for three minutes without notes definitely does.

It is the output children get least practice at. School is heavily weighted toward writing. Presentation happens a few times a year, under pressure, in front of peers. Low-stakes speaking practice is genuinely scarce.

Editing teaches structure backwards. Listening to your own recording and deciding what to cut is a fast, brutal, effective lesson in what was actually interesting.

And AI's role here is deliberately small. It helps write the questions. It does not host, script, or narrate. If it did, the child would be reading rather than speaking — and reading aloud is a completely different (and much less useful) exercise.

What you need

Three formats that work

Pick one before starting. They have quite different difficulty levels.

Format A · The expert episode (easiest, best first choice)
Your child talks about something they already know a lot about — a game, an animal, a hobby. You are the interviewer asking curious questions.

Format B · The interview (best material)
Your child interviews a real person: a grandparent, a neighbour, a parent about their job. The AI helps prepare the questions.

Format C · The investigation (hardest, best for 11-12)
Your child researches a question and reports what they found, with an opinion at the end.

Start with A unless there is an obvious interview subject available. A is lower stakes and teaches the format; B produces better audio but needs a willing participant.

The shape