Entry Points Into Machine Learning

In today's episode, we'll explore three tools you can use to introduce Machine Learning to your students: Google's Teachable Machine, Micro:bit CreateAI, and Grammarly's Complete Guide to Machine Learning. Visit AVID Open Access to learn more.

Transcript

Paul Beckermann 0:00 Welcome to Tech Talk for Teachers. I'm your host, Paul Beckermann. Check it out.

Transition Music with Rena's Children 0:06 Check it out. Check it out. What's in the toolkit? Check it out.

Paul Beckermann 0:16 The topic of today's episode is entry points into machine learning. Over the past three years, teachers have gotten a crash course into generative AI tools like ChatGPT, Gemini school AI, and Magic School. While these front-facing application tools have gotten demystified a little bit, the underlying science of machine learning remains more of a mystery to many. It sounds scary, really complicated, or maybe even completely off the radar.

To get us all on the same page, let me define machine learning according to the Oxford English Dictionary: machine learning is the use and development of computer systems that are able to learn and adapt without following explicit instructions. They do this by using algorithms and statistical models to analyze and draw inferences from patterns in data.

Essentially, machine learning is the science that underlies artificial intelligence applications like ChatGPT and Grammarly. In fact, Grammarly offers its own definition of machine learning on its website, writing: "Machine learning, or ML for short, is a subset of AI that enables computers to learn from data, rather than through explicit programming, by recognizing patterns in data. ML allows computers to make predictions and improve over time. This capability powers technologies such as self-driving cars, streaming recommendations, and Grammarly real-time writing suggestions."

While I admit, teachers don't need to become machine learning experts to use AI, it can be helpful to have a basic understanding of how it works. This understanding can help make sense out of the responses you get from AI applications. It demystifies the experience and can help us remember that while these machines are learning, they're not really thinking on their own. They're following algorithms—computer programs—to predict patterns based on what they've been taught.

This awareness is not only helpful for you as a teacher, but it can be extremely valuable to your students. After all, they're growing up in a world saturated by machine learning and AI, and we can do them a great service by empowering them with an understanding of this powerful technology. Some may pursue a career in technology, and those that don't will certainly be living their daily lives interacting with it. So in that light, I'd like to share three entry points into machine learning that you can use yourself and also introduce to your students.

Transition Music with Rena's Children 2:48 Here are your three here are your free tips. Here are your three tips.

Paul Beckermann 2:54 Number one: Teachable Machine from Google. This is a browser-based tool that you and your students can use to train a computer to recognize things like images, sounds, or poses. It's free, easy to use, and doesn't require any coding.

Essentially, users get to train the Teachable Machine to recognize a type of image, a sound, or a pose. For images, you could train it to distinguish between a picture of a person or a dog. For audio, you might teach it to know the difference between snapping your fingers or clapping. For poses, it will use your webcam to recognize body position, so you can teach it to recognize a movement like squatting or pointing. To use it, you click on one of these choices, and then the website walks you through the process. For example, I clicked "audio project" and was taken to a screen where I could either record or upload different audio clips. It began by prompting me to record 20 seconds of background noise.

Paul Beckermann 4:02 Then I labeled and recorded eight or more samples of a certain type of sound. I chose to use clapping for the first one, and then I added a second class called finger snapping, and recorded eight more samples of those. Once my samples were recorded, I clicked on "train model." After about 10 or 15 seconds, it was done, and I could try it out on that same screen without going anywhere else. An output meter at the bottom showed me what type of sound it thought I was creating, and showed me its level of confidence in that response by showing a percentage number.

I did this twice, and found that my first training appeared to be more effective than the second. This could lead to some really great discussions about what impacts effective training inputs have on the machine learning. If I liked what I created, it allowed me to download or embed the project with some code. Google's Teachable Machine could be a great way to introduce students to machine learning in a simple way.

Paul Beckermann 5:07 You might even find curricular ties for the experience, like categorizing types of leaves in science class or different physical movements in physical education.

Number two is Microbit Create AI from Microbit. This is another free browser-based tool. However, this one is intended to be used in conjunction with the microbit device, which you would need to purchase. The microbit device is a pocket-sized computer with built-in sensors such as an accelerometer, a compass, and a microphone. For this application, students would wear it on their wrist or ankle to track movements that they make.

Similar to Google's Teachable Machine, students use the website to train the microbit to recognize certain movements, like the difference between waving and clapping. Again, students record multiple samples of each movement to train the computer. Each of the trained movements is reflected by a different LED shape showing up on the microbit itself.

Paul Beckermann 6:11 This offers really great immediate feedback to the students. Students can also see the difference in movement by studying an accelerometer graph on the website. Students will likely see that the recorded wave patterns begin to represent their specific movements. They'll see trends. If the results don't seem accurate, students can retrain their microbit. Again, it's very similar to the process used by Google's Teachable Machine. Finished models can be edited in MakeCode and exported to the microbit so it can actually run without a computer attached, using simple block coding in MakeCode. Students can go deeper yet, if they want to, and they can create more advanced machine learning projects.

Number three: Your Complete Guide to Machine Learning from Grammarly. This third option is a resource rather than a tool.

Paul Beckermann 7:06 While there are lots of materials online that you can use to learn about machine learning, this webpage from Grammarly is a great place to start if you simply want to learn more about machine learning and better understand what is involved. It's also a great way to understand how editing tools like Grammarly work and how they are impacting your own writing based on the machine learning used to train them.

This page is broken down into categories of information: types of machine learning, common machine learning tasks, machine learning techniques, and machine learning concepts. The level of writing is more appropriate for a secondary level student. So if you're an elementary teacher, it's probably best to use it as a way to gain understanding before sharing it with your students in more age-appropriate language.

Paul Beckermann 8:01 If you're a secondary teacher, you could assign students to review this material on their own, either before or after experiencing machine learning with either Teachable Machine or Create AI.

So in general, Teachable Machine, Create AI, and the Grammarly guide each offer a unique entry point into the world of machine learning, whether it's interactive, hands-on, or conceptual. By introducing students to these tools, we can teach them how AI works, while also giving them the skills to think critically about the technology shaping their world.

To learn more about today's topic and explore other free resources, visit Avidopenaccess.org. Specifically, I encourage you to check out the article collection, "AI in the K-12 classroom," and of course, be sure to join Rena, Winston, and me every Wednesday for our full-length podcast, Unpacking Education, where we're joined by exceptional guests and explore education topics that are important to you.

Paul Beckermann 9:11 Thanks for listening, take care, and thanks for all you do. You make a difference.