It’s always fun starting a topic with the history and a timeline is easy and a great discussion starter. This timeline by the Queensland Brain Institute is clear and leads the students to ask “why are there decades when nothing happened?”
Machine Learning is a way to achieve Artificial Intelligence. In machine learning, we teach the machine by training it with lots of examples of data demonstrating what we would like it to do so that the machine can figure out how to do it on its own.
There are plenty of apps and software to use to engage students in appreciating and understanding how this works. Here are some idea:
Aipoly Vision app to identify objects around the classroom.
“Seek by iNaturalist” is a child-friendly version of iNaturalist built on the same dataset but designed for outdoor environments or around the school.
Look at Picasso’s line drawings, where he uses just a few lines to craft an image. Then try out Quick Draw, to demonstrate how AI is using knowledge of shapes, pixels and feature extraction to guess what a human is drawing.
Use PixelArt to explore drawing images through colouring individual pixels, or do this with a pencil and graph paper! Created by my students…
Use Edge Detection to see how the computer would do it!
Fabulous activity by Cognimates provides the platform for students to train AI models. It uses computer vision. There are a few projects available for students try. These projects are using a supervised learning approach.
Cognimates: a visual programming platform build as a Scratch extension by MIT Media Lab (creator – Stefania Druga). This is a free platform for everyone.
Clarifai vision API: Cognimates platform uses Clarifai vision API to build custom AI projects.
Machines learn through supervised and unsupervised learning approaches. Supervised learning happens when humans label examples of data which helps the machine to classify….. eventually on its own, whereas unsupervised is when the machine detects patterns from large amounts of data and makes its own decisions.
Natural Language Processing (NLP) is the ability for machines to interpret and analyse forms of human communication, such as text and speech.There are a number of NLP tools available, such TwinWord, that can be used to explore the relationships between words.
Using the game in Google Experiment called Semantris students can see how AI is using NLP datasets to work out best word associations.
Safari Mixer lets the student invent a brand-new animal by combining the head, body, and legs of other animals. Simply tell Safari Mixer which animals you want to combine, then see (and hear) your creation come to life. Using AI, Safari Mixer gives a complete audio and visual experience of the student’s unique creation.
Students can make a project using text, click on this link Machine Learning for Kids and select a project to get started. There are plenty of projects for both images and text modelling.
There are also plenty of unplugged activities to enjoy which demonstrate how machines learn:
How do we learn? (rote, discovery, play) If you have a pet, teach it a new trick.
Play classification games, organise the classroom, building blocks into size, colour…
Memory games of turning over pictures which match. I Spy Game with attributes, describe objects to another student see how many guesses it takes.
Play Pictionary to understand how computers look for lines and shapes.
Play AI Bingo Game which demonstrates how much AI is around us. Print Bingo cards and question sheet.
Family Feud is also another good game which demonstrates the collection of data. The link will give questions and data for games within the classroom.
These ideas provide plenty of opportunities to have some fun while learning more about AI and Machine Learning. Previous post is here
This ‘Age of AI’ series on You Tube is also excellent.