Uma casa em 3D, um pequeno filme ou um jogo de tabuleiro: comecem com um projeto de IA em família grátis. Conhecer a adesão →

Code.org AI for Oceans

Label ocean pictures and see how training examples shape a classifier.

Free
🟢 Low Risk
Ages 8-14

Catalog listing updated November 2026

Visit Website

1-Tap Parent Pulse: Where do you stand on Code.org AI for Oceans?

Tap the option that matches your question right now to unlock a practical family setup tip.

Get the parent newsletter and a free 10-point checklist to review Code.org AI for Oceans’s age and privacy rules

Open checklist now · Open the free homework agreement

Parent email only · No child data · Unsubscribe in 1 click

This free activity uses supplied pictures to train a classifier and try it on new examples. The choice of examples and labels can produce bias that children can discuss.

Funcionalidades principais

  • Image labeling
  • Classifier predictions
  • Bias discussion

Exemplos de utilização

  • 1.Distinguish fish from debris
  • 2.Record one misclassification
  • 3.Compare training sets

Vantagens

  • ✓No login needed
  • ✓No photo uploads required
  • ✓Errors provide discussion material

Limitações

  • ✗Narrow example domain
  • ✗Limited model internals
  • ✗Short activity needs follow-up

Parent Setup & Supervision Guide

Agree on what counts as a fish. When the model makes a mistake, identify a missing kind of example and compare the next round.

Frequently Asked Questions about Code.org AI for Oceans

Do we upload family photos?

Use the pictures already supplied in the activity.

What can a wrong prediction tell us?

Check for missing examples and inconsistencies in the labeling rule.

Projects Using This Tool

No projects are using this tool yet

Check back soon as new projects are added.

Share:

Tool Information

Pricing

Free

Platforms

web

Languages

English
Chinese

Ages

Ages 8 - 14

Categories

AI learning
AI literacy
Coding

Risk level

🟢 Low Risk

No account or personal training images are needed for the activity. Explain that a model can repeat a labeler's bias.

Added

11/21/2026