Learning "withGoogle" Favorite Resources

  • Published: Monday, June 22, 2020

A few years ago, a teacher approached me. She said, "I know how to make a Doc. I use Slides and Forms and Classroom every day. What else is out there?" I had been so focused on teaching the basics, her question really gave me pause. What else IS out there? So, I started doing some research on just that. What were the other things that Google had for educators? The sheer volume of information I got back astounded me. Besides full programs like Applied Digital Skills and Be internet Awesome, which deserve their own blog posts, I found that Google is neck deep in artificial intelligence and machine learning. Moreover, they make many of their “experiments” available to the public. Some of these are fantastic, both as resources for the classroom and as ways to introduce students to these two huge and complicated topics. There are more than 1500 experiments to choose from! But I have pared it down to a list of my five favorites.

Quick Draw: Ever had a computer guess what you were drawing and then tell you what it is? That is what Quick Draw is. You have a limited amount of time to draw a suggested object. As you draw, Quick Draw will continually speak its guesses (some of which are hilarious). Until it finally is sure what you are drawing and then it will say "Oh I know. It’s…" And it is pretty good at guessing. This is an experiment in machine learning. It has been gathering a dataset from all the players’ drawings over time and it is learning what those shapes actually are.

Talk to Books: Learning to use a search engine well is a critical skill. But search engines have limits. They can’t search for moods, or feelings or other, more subjective, content. Until Talk to Books came along. You simply ask it a question such as "What is it like to live in the mountains" or "How does it feel to float in space?" It then searches the library at Google Books for passages from books that describe your question. It then returns the book with all its information. Talk to Books is a great tool for helping students build empathy or research creative writing projects for which they have no frame of reference (e.g., asking students in Missouri what it is like to live by the ocean).

Teachable Machine: This is a very simple, easy to use, machine learning experiment. You "teach" the computer what two objects are by showing them to the camera and identifying what you are showing. Then when you run the scenario, the machine will "estimate" the likelihood of what it is seeing in a simple bar graph below the screen.

AutoDraw: This is kind of a combination of Quick Draw and Teachable Machine. Google hired several artists to build a band of simple graphics for all kinds of things. Let’s say you need a graphic of a simple coffee cup. Go to AutoDraw, turn it on and start drawing. Based on what it sees (or thinks it sees), AutoDraw will begin to populate the top of your screen with graphics it thinks you are trying to draw! If you find one you like, just click it and AutoDraw will swap out your drawing for the graphic. Then just download it and use it!

Ngram: My final favorite is more of a big-data visualizer. It allows you to compare related (or unrelated) topics in graph form. It is better for historical data than current events, but you just type in your search terms and date ranges, Ngram will then search Google Books and return all instances of those search terms mentioned in books over time. Since literature usually reflects current events of the day, typing in "wagon" vs. "automobile," for example, shows an interesting graph of when literature started talking about autos, how long it took for autos to be discussed more than wagons, when automobile discussions peaked, and more.