Build a Teachable Machine Image Classifier With Your Kid

Introduce your child to the fundamentals of machine learning through a hands-on, zero-code computer vision project.

In an era where artificial intelligence is reshaping every industry, the most valuable skill we can teach the next generation is not just how to use AI tools, but how to understand how they think. For busy executives, finding high-quality, educational activities to share with your children can be a challenge. Many coding kits require expensive subscriptions, complex setups, or hours of frustrating troubleshooting. This weekend, you can bypass the complexity and build a real, functioning machine learning model with your child in less than thirty minutes. Using a free, web-based tool developed by Google, children aged 8 to 16 can experience the entire lifecycle of AI development: from data collection and model training to testing and deployment. This hands-on project demystifies artificial intelligence, transforming it from a magical black box into a practical tool that responds to data. By building an image classifier together, you will give your child a foundational understanding of computer vision while spending meaningful, collaborative time together.

Step 1: Setting Up Your AI Lab

To begin this project, you do not need any specialized hardware, paid software, or programming experience. All you need is a computer with a working webcam and an internet connection. Sit down with your child and open a web browser. Navigate directly to the official Google Teachable Machine website at https://teachablemachine.withgoogle.com/ and click on the "Get Started" button. On the project selection screen, choose "Image Project" and then select "Standard Image Model". Explain to your child that you are going to teach the computer how to recognize two different objects, just like a human brain learns to identify things. Let your child choose the two objects they want to use. Excellent choices include a favorite toy versus a book, a coffee mug versus a water bottle, or even simple hand gestures like a thumbs-up versus a thumbs-down. This choice gives your child immediate ownership over the project and sparks their curiosity from the very beginning.

Step 2: Gathering Your Training Data

Once your project is open, you will see two default boxes labeled "Class 1" and "Class 2". These represent the categories your AI will learn to recognize. Have your child rename these classes to match the chosen objects (for example, "Action Figure" and "Notebook"). Now, it is time to collect the training data. Click the "Webcam" button under the first class. Instruct your child to hold the first object in front of the camera. Click and hold the "Record" button to capture images. As you record, have your child slowly rotate the object, move it closer to and further from the camera, and change the angle. Explain that the AI needs to see the object from many different perspectives to truly understand it. Aim to capture at least one hundred images. Repeat this exact process for the second class using the second object. This step teaches children that AI models are only as good as the data we feed them.

Step 3: Training Your Custom Model

With your training data collected, you are ready for the magic step: training the model. In the middle column of the screen, you will see a box labeled "Training" with a blue button that says "Train Model". Have your child click this button. It is crucial to instruct them to leave the browser tab open and untouched while the training process occurs. The training takes about ten to thirty seconds as the algorithm analyzes the patterns, colors, and shapes in the images you captured. Use this brief waiting period to explain what is happening under the hood. Tell your child that the computer is looking for common features in the images (such as the roundness of a mug handle or the bright colors of a toy) and building a mathematical rulebook to tell them apart. You can explain that this is similar to how they practice a new skill: each round of training is like a practice session.

Step 4: Testing and Iterating

Once the training is complete, the "Preview" column on the right side of the screen will activate, turning on your webcam. Now comes the most exciting part of the project: testing the model. Have your child hold up one of the objects in front of the camera. The model will display a real-time confidence rating (from 0% to 100%) showing which object it thinks it is seeing. Challenge your child to trick the model. What happens if they hold the object upside down? What happens if they only show a tiny corner of it? What happens if they hold up a completely different object? If the model makes a mistake, explain that this is a normal part of AI development. Go back to the training classes, record more images from the angles where the model struggled, and click "Train Model" again. This iterative process teaches kids that building AI is a continuous cycle of testing and refinement.

Parent and Educator Sidebar

Core AI Concept: Supervised Learning This project demonstrates supervised learning, where an AI model is trained on labeled data (images that we explicitly named "Action Figure" or "Notebook"). The model learns by finding patterns in these examples so it can make predictions on new, unseen images.

Conversation Starters:

• "How do you think a self-driving car uses this kind of technology when it is driving down a busy street?"
• "If we trained our model using only red toys, do you think it would recognize a blue toy? Why or why not?"
• "Why is it important to make sure we give the computer high-quality, diverse pictures when we are training it?"

Action Steps Summary

• Open a web browser and navigate to https://teachablemachine.withgoogle.com/ to start a standard image project.
• Select two distinct objects or hand gestures to use as your training classes.
• Use your computer webcam to capture at least one hundred diverse images for each class.
• Click the Train Model button and keep the browser tab active while the computer processes the data.
• Test the model in the preview window, identify any errors, and retrain the model with better data to improve accuracy.

Ready to inspire the next generation of tech leaders?

Block out 30 minutes this Saturday morning to build this classifier with your child.