Overview
Welcome to Week 9 of Full STEAM Ahead! This week, we’ll be learning all about artificial intelligence 🤖💻.
Have you ever watched a video that YouTube recommended to you? Have you ever used a Snapchat filter to add dog ears to your head? Have you ever asked Alexa or Siri a question? If so, you have used artificial intelligence! Artificial intelligence, or AI, is a computer system that performs tasks such as visual perception, speech recognition, and decision-making. It not only powers some of the technologies that we are familiar with, but is used in fields such as healthcare and space exploration 🏥🚀.
In this package, you’ll get the chance to learn about and create systems that include artificial intelligence. These activities were designed by the Personal Robots Group and the App Inventor teams, which are both a part of the Democratizing AI Initiative at MIT. Through these activities, you’ll get the chance to:
- Identify AI systems that you use every day 🔍
- Think deeply about the data that powers AI technologies 📈
- Design AI systems that take multiple perspectives into account 🎨
- Create your own apps that contain models that recognize images and audio 👩💻
If you enjoy the materials here, please check out some of our other resources and materials at aieducation.mit.edu.
Make sure you scroll to the bottom for additional activities.
Activities
| Introduction to AI | |
Do you know what artificial intelligence is? It’s a technology that is all around us. Most of us have it in our homes, and we use it every day. There are several activities in this document that will help you to define AI, learn about chatbots, and make your own image classifier! | |
| Mystery YouTube Viewer: A Lesson on Data Privacy | |
In this activity, we will observe a YouTube home page from a mystery user. Using the clues from the videos this character has been recommended, we will ask you to draw what the character might look like, guess what they might like to do, and figure out where they might live. What you will do is similar to what YouTube algorithms can do: make guesses about you based on what you watch! a. Video: Overview Video
| |
| AI Bingo | |
Students are given bingo cards with various AI systems. Students find a partner who has also used that AI system and together work to identify what prediction the system is making and the dataset it uses. | |
Activities
| Introduction to Algorithms As Opinions | |
Students learn that algorithms, like recipes, are a set of instructions that modify an input to produce an output. Students are then asked to write an algorithm to make the ”best” peanut butter and jelly sandwich. Students then explore what it means to be ”best” and see how their opinions are reflected in their algorithms. | |
| Introduction to Supervised Machine Learning and Algorithmic Bias | |
Students are introduced to the concept of classification. By exploring Google’s Teachable Machine tool, students learn about supervised machine learning. Then students are asked to build a cat-dog classifier but are unknowingly given a biased dataset. When the classifier works better on cats than dogs, students have the opportunity to retrain their classifiers with their own new datasets. | |
| Strike a Pose! | |
Students learn to use the image recognition and pose recognition branches of Google’s Teachable Machine, training a supervised machine learning model to classify with given labels and testing out the strengths and weaknesses of different representations of data. By training an image and pose model to recognize the same set of body movements, students explore how data representation affects a model’s learning and performance. | |
| YouTube Redesign | |
Students construct an ethical matrix around the YouTube Recommender Algorithm. Based on this ethical matrix, students determine a goal (or “opinion”) for their algorithm. Students then paper prototype what this new version of YouTube would look like and imagine features that meet the values their identified stakeholders have. | |
These activities will show you how to make mobile apps that use artificial intelligence in some way. To lay the groundwork, read this set of interviews by YR Media, titled “In the Black Mirror”.
You might also listen to this YR Media podcast, Robots Taking Over This Ish, with high school seniors talking about privacy in the age of surveillance, even by one’s own parents!
For the following activities, you will be using MIT App Inventor, software that runs in your browser, and helps you create apps that run on Android devices. If you haven’t used App Inventor before, we suggest you try a couple of our introductory tutorials, to become familiar with the software and the coding blocks.
You will need a Gmail account to log into the MIT App Inventor coding environment at http://ai2.appinventor.mit.edu. Any of the activities/tutorials below may be completed using our alternate server, http://code.appinventor.mit.edu. You will not need a Gmail account to login at this site, but your projects are not permanently saved.
Activities
| Introduction to Machine Learning & Image Classification: Make the WhatIsIt App (Beginner) | |
Learn the basics of machine learning and create an app that identifies objects! This app uses a pre-trained model, just to give you a taste of how image classification works. Note: To make this app and try it out on a mobile device, you should first check that your device has the hardware capability to run it. Part 1 – Learn about Machine Learning and try out Google’s Teachable Machine Part 2 – Make the WhatIsIt App Teachers: full lesson plan here. | |
| Expression Match Game with Image Classification (Beginner) | |
As a followup to the What Is It app, learn to train your own image classifier by posing with different facial expressions. Then build a game app where users have to match the expression. Can you make a happy face? A sad face? A scary face? Note: To make this app and try it out on a mobile device, you should first check that your device has the hardware capability to run it. Part 1 – Train your model using the Personal Image Classifier Part 2 – Create the Expression Match Game App
Teachers: full lesson plan here. | |
| Voice Authentication Diary App with Personal Audio Classifier (Beginner) | |
Train an audio classification model to detect your voice saying “hello” to unlock your personal diary app. Note: Before making this app, download this apk file and test that it runs on your mobile device. | |
| Therapist Bot Tutorial (Intermediate) | |
Make your own version of ELIZA, the famous computer program that simulates conversation, this time in mobile app form.
- Complete the tutorial on the MIT App Inventor site (requires a Gmail account)
- Complete the tutorial on the Code App Inventor site (anonymous login, but projects are not permanently saved.
| Rock, Paper, Scissors Tutorial (Advanced) | |
Make a mobile app that trains itself to beat its opponents in the famous game, Rock, Paper, Scissors.
- Complete the tutorial on the MIT App Inventor site (requires a Gmail account)
- Complete the tutorial on the Code App Inventor site (anonymous login, but projects are not permanently saved)