Welcome to the Teachable Machine Unit!
In this unit you will learn how to train a real AI model using Google's Teachable Machine — no coding needed! You'll create image classifiers, explore bias, test accuracy, and design your own AI product idea.
What You Will Learn
- How machine learning models are trained using labelled data
- How to create a 2-class and 3-class image classification model
- What bias means in AI and why it matters
- How to test a model properly using separate training and testing data
- How to calculate accuracy and improve a model
- How to design an AI product idea and evaluate it
- How to present your work and reflect on what AI can and cannot do
What You Will Need
Equipment for every lesson
- A Windows PC with a web browser (Chrome or Edge)
- Microsoft Paint (for drawing shapes)
- A folder on the desktop to save your images
- Access to teachablemachine.withgoogle.com
Note: No camera or microphone is needed! You will use drawn images, not photos.
Your Progress
Safety & Expectations
Classroom Rules for This Unit
- Only use Teachable Machine for the tasks your teacher sets.
- Do not take photos of other students or upload personal images.
- Save all your work in your designated folder.
- Be respectful when discussing AI — listen to others' ideas.
- Ask your teacher if you are stuck — it's OK not to know!
How to Use This Site
- Use the sidebar on the left to navigate between lessons and pages.
- Each lesson page has step-by-step instructions — follow them in order.
- Fill in tables and answer questions directly on the page — your work is saved automatically.
- When you finish a lesson, click the green "Mark Lesson as Done" button.
- Use the Resources page to download worksheets, help sheets, and datasets.
- If you need help, check the Help Sheet page or ask your teacher.
Unit Overview: This is a 6-lesson unit designed for Primary School students. Each lesson is 45–60 minutes. Students use Google Teachable Machine to build image classifiers using hand-drawn shapes (no camera/mic needed). The unit covers: building models, bias, testing, multi-class classification, product design, and reflection/ethics.
All worksheets, starter slides, and datasets can be downloaded from the Resources page.