How Jubbo works

Very little, deliberately. Jubbo has a microphone, a speaker, two round displays for the eyes, and one switch, for power. It has no camera, no screen you could watch, and no way to browse anything.

It also has no intelligence of its own. Everything it says is decided on our servers and sent to it as sound — the box is a speaker and a microphone with a good internet connection. That is why it needs Wi-Fi to do anything at all, and it is also why we can change what it does without you ever updating it.

Start an activity from the app. The eyes change to show Jubbo is listening, and the microphone opens — it is closed at every other moment.

Your child speaks. When they stop, the microphone closes and the eyes show it is thinking; there is a soft tick while it does, because a silent box makes a child wonder whether it heard them. Then it answers, and either asks the next thing or ends the round.

You can end it at any point from the app. There is no wake word, and nothing you say switches it on.

The eyes are how Jubbo shows what it is doing — listening, thinking, pleased that an answer was right. They are expression, not personality: there is no character behind them who remembers your child or misses them when the box is off.

Nothing. Between activities Jubbo rests. The microphone is shut, and at most it will say one short line every ten minutes to show it is still on. It will not start a conversation, and it cannot.