Skip to main content

Now inviting primary schools across the UK to host Junior AI Explorers clubs. Host a club →

Home: AI ExplorersAcademy

Ages 8–11 · After-school AI clubs · Partnering with UK primary schools

Try it at home

Simulation · no data collected

Can you teach a computer to save the ocean?

Sort sea creatures from rubbish, train a simple model and see what it gets right — and wrong. It's a taste of Mission 01, perfect for doing together.

1Label the training data

0/6

Is each one a sea creature or rubbish? Your labels are the examples the model learns from. (Try labelling one wrongly later and see what happens!)

  • Clownfish
  • Jellyfish
  • Crab
  • Plastic bottle
  • Drinks can
  • Carrier bag

Label at least one sea creature and one piece of rubbish to train.

2Test on new items

Label the training examples, then press “Train the model” to see how it sorts things it has never seen before.

How this simulation works

This is a simple simulation, not a real trained AI model. Each item is described by five features we chose (has eyes, has fins or legs, is shiny, is soft, has straight edges). When you test a new item, the program finds the most similar labelled example and copies its label — a method called “nearest neighbour”. Real machine-learning models work with far more data and features, but they share the same big ideas: they learn from examples, they're only as good as their data, and they can be confidently wrong. Everything runs in your browser; nothing you click is stored or sent anywhere.

Enjoyed that? There are six missions in the full course.

Junior AI Explorers is a six-week after-school club for ages 8–11.

Explore the course