ML for Kids

Machine Learning for Kids: helping kids learn about ML by letting them train models and make things with them Info at

Make a AI-powered traffic camera game in #Scratch, where you have to train a #machinelearning model to recognise if there are passengers in the car. Full instructions at

A new feature was added to @MLforKids this week by @all_about_code - integration with EduBlocks! 🎉🎊 This will let students create machine learning Python projects in the browser, by dragging and dropping blocks on a canvas.

RT @dalelane: The recording from my talk at @Heapconf has been posted - the idea of the talk was to share stories of #artificialintelliegence lessons I’ve run in schools. The focus of the talk was how I've seen children understand and react to #machinelearning

RT @cintichuEF: @chusca14 @sandboxeduca @rafabillor @genially_es Estoy haciendo un curso de: DESCUBRE LA INTELIGENCIA ARTIFICIAL IBM: Machine Learning for Kids. También incluye Scratch 3. El curso nos lo ofertan desde la plataforma de innovación de la comunidad de Madrid.

Creating projects based on real-world applications of #machinelearning is an effective way to open up students' eyes to the fact that AI systems are all around them, and impact their lives in countless ways

RT @RaspberryPi_org: EXCITING NEWS 🎉 Experience AI launches today in partnership with @DeepMind. Our new AI and machine learning programme for teachers, students, and other educators. Find out more 👉 #AI #MachineLearning #DeepMind #ExperienceAI

Make a #machinelearning-powered "I Spy" game in #Scratch, where you have to guess what the AI has recognised in a photo Full instructions at

School students are able to engage in debates about the ways that artificial intelligence tech is used. This can be helped giving them a hands-on chance to experiment with the tech and see for themselves how bias can be introduced.

Some of the most effective #machinelearning lessons are when something goes a bit wrong. The discussions this enables really helps students understand how ML behaves. Here's one example of how that can go...

Allowing students to experiment in a child-friendly sandbox can help them to understand how #MachineLearning doesn't always learn what we wanted it to learn! Here are a couple examples of how I've seen this go...

The flexibility of #Scratch provides students with opportunities to explore the behaviour of #MachineLearning models in creative ways - such as how to deal with machine learning tech being a black box.

Students can learn a lot about the role of confidence scores through playing and experimenting with their own simple #machinelearning models - some of the ways children will describe this are fantastic

The experiences students get by making their own #machinelearning projects helps them understand the behaviours and motivations of big tech companies. This is one of my favourite examples:

Sentiment analysis projects are a nice way to show students that machine learning can learn to recognize a variety of types of patterns in text, not just the meaning or intent of the text.

Making simple #machinelearning projects introduces students to the idea of keeping training and test data separate

letting students try each other's #machinelearning projects allows them to notice for themselves how ML models trained with a wider variety of training examples perform better

Students can see for themselves why crowd-sourcing and gamification are used to help create training data for #machinelearning projects

Giving students freedom to experiment with their own machine learning model lets them learn for themselves the relationship between the quantity of training data and the accuracy of machine learning models.

Going through the process of creating a #machinelearning project for themselves gives students an insight into how these systems are created in the real world.

Using existing tools like #Scratch and #AppInventor to introduce students to #machinelearning was a conscious choice - to send a message that it complements what they already learn about coding