Machine Learning for Kids
Authentic AI-powered projects make kids' creations feel like magic
Take a look inside 10 images
Machine Learning for Kids
Pros: Fun and exciting projects illustrate exactly how machine learning is being applied in the real world.
Cons: Setup takes some time and patience.
Bottom Line: It takes a topic that would be difficult to teach in any practical way and makes it accessible for the classroom.
For any teacher who wants to introduce students to artificial intelligence and machine learning, Machine Learning for Kids is likely your new best friend. It will take some time to set up and prepare, but once you've done it, you'll have an amazing resource that illustrates exactly how machine learning is being applied in the real world. To start, you'll sign up for an account with the site and then create an account with IBM Cloud in order to access Watson. From there, you'll use the worksheets (which are actually very clear sets of instructions) to make each project. In the end, you will have trained a simple Scratch program using your own data set and Watson. The site also works with Python and App Inventor.
Machine Learning for Kids is fairly complex, so you won't be able to just set your students loose on the site and expect them to come away understanding AI. You should take time to read how the site works, choose a project, and follow the instructions from beginning to end yourself before leading students through it. Having students work in small collaborative groups is a great way to make the experience more fun and to have students troubleshoot any problems. The people who run Machine Learning for Kids provide detailed tips for any problems you might run into and also personally answer questions incredibly quickly.
Machine Learning for Kids is a website offering projects that combine machine learning and artificial intelligence (AI) with Scratch (the block-based coding language) and App Inventor. Users sign up for a free account from IBM Cloud to get access to Watson, the question-answering computer system. Machine Learning for Kids then provides step-by-step instructions for creating a data set that's "fed" to Watson to train the AI. Through a simple Scratch interface, students can then see the machine learning in action.
Currently, there are 27 different projects available, all of varying types and complexity. Each one makes use of machine learning in a different way to show the ways in which it's being used in research, business, and academia. Projects include:
- Creating a chatbot that recognizes questions and gives plausible answers
- Using the computer's camera to sort photos
- Facial recognition
- An app that makes recommendations based on user preferences
Machine learning and artificial intelligence play increasingly large roles in our lives whether we like it or not (and, indeed, whether we even know it or not). Everyone should have an understanding of how it works simply because it's shaping our lives. Without a tool like Machine Learning for Kids, it would be very difficult to teach about it in any practical way, and the creators of the site have done an incredible job of making it fun and accessible. It's not easy, but really good learning seldom is.
Students don't need a lot of experience with Scratch to use Machine Learning with Kids, because the site provides the code. Understanding how Scratch works will make it easier to recognize the role AI is playing in the end results, but you won't have to teach your students Scratch in order to use the site. This allows you and your students to focus on training the AI and improving your data set. It also lets you focus on conceptual understanding and analysis of the potential benefits and shortcomings of machine learning.
Though it may not be simple at first, any difficulty you have in setting up Machine Learning for Kids is absolutely worth it as soon as you see the reaction on your students' faces upon seeing their AI-powered projects run. You'll probably be just as excited as them.
Learning Rating
Overall Rating
Explore Our Favorite Tools
