Build Your Own Machine Learning Games

About This Class

Your child builds a video game that gets smarter the more it plays, powered by a machine learning model they train themselves. They run the full loop real AI engineers use: collect data, train, test, and wire it into a game. It is the difference between using AI and building it.

Every session features hands-on projects where kids train models to recognize hand gestures, keypresses, or voice commands, then connect those models directly to games they build in Scratch. Along the way, students learn about dataset size, accuracy, bias, and the engineering mindset.

What Makes This Class Different

• Train real machine learning models using images, text, and sounds
• Integrate custom AI models directly into Scratch games
• Learn the entire machine learning pipeline from data collection to testing
• Get personalized guidance with small class sizes

Skills Your Child Will Learn

• Understand core machine learning concepts like training data and model accuracy
• Collect and clean data to train image, text, and sound classifiers
• Program game mechanics in Scratch that react to custom AI inputs
• Identify and resolve bias and data issues in machine learning models
• Apply the design loop to test and improve AI-powered projects

Parent Participation

Parents are welcome to watch and check out what their child builds, and we encourage playing the games together at home.