Start with one of these

A family in Amman buys a new television, and the remote has a small button with a microphone printed on it. Press it, speak, and the set is supposed to do what you ask.

The aunt who lives with them loves the idea at once. She holds the button and talks to it the way she talks to her nephews: put on the channel with the old films, the one from last night, and make it a bit louder because the kitchen is noisy. Nothing happens, or the wrong thing does. By the third evening she has stopped pressing it.

The remote understands a fair amount. It does not understand that. What it needed was not a cleverer microphone but a small card on the back, listing a dozen things it handles well. Say the name of a channel. Say louder or quieter. Say the name of a film.

This happens with almost anything new that seems to listen. People see a little capability and imagine a great deal more. They speak to it as they would to a person, and when it fails them, they decide it does not work at all. The truth sits somewhere in between, and nobody has shown them where.

A handful of examples fixes that gently. It does not stop anyone trying something else. It teaches the outline of what is possible, so the first few attempts succeed, and early success is what brings people back for a fourth.

I think this is worth remembering whenever the people we serve hesitate in front of something we have made. The reflex is to explain all of it, or to insist it can do anything. The more useful offer is usually smaller: here are some good things to ask for, start with one of these.

Most confusion is not a shortage of intelligence. It is a shortage of a starting point.