This episode features an interview with the authors of a new book from Make: titled “AI Robots” which includes Reade Richard, Andy Forest, Brenda Shivanandan and Denzel Edwards. The book teaches kids how to build AI-powered projects using hands-on activities and coding. Each project has four different sections that include the physical build, electronics, coding with Micro:bit, and adding AI capabilities. They emphasize the importance of integrating different subjects, such as STEM, in a project-based way. The book aims to empower students and parents to explore AI technology and learn how to use it responsibly.
Transcript
Dale: Make has just published a new book titled AI Robots. The team behind the book are my guests on this episode of MakeCast. AI Robots is a book full of hands on projects where kids can build things that use AI as a visual recognition system.
Each project in the book has four sections. The first is a physical build, the second is adding electronics, and the third is writing some code for Micro:bit, and the last section is using AI to add a special feature. Before we begin, I’d like to show a few pages from a sample chapter and describe them for those that are listening.
This second chapter, cars and axles, is a challenge to build a simple car out of cardboard, add motors to it, write code for a micro bit, and then add AI to help the car respond to its surroundings.
In the first section, we see the information on how to build the physical car, and we end up making it more than a car by decorating it as an ice cream truck.
Then in the electronics section, we add a motor to be able to power the car.
Next, in the coding section, we use MakeCode to write the code for Microbit that plays ice cream truck music when a button is pressed.
Finally, in the fourth part, we get to add AI to the project, and we use a digital camera to take pictures of buildings and other things that train the AI’s visual recognition system, and then we write code to respond to what the AI recognizes.
These are pretty cool projects, and I think, teachers, parents, and kids will have fun doing them.
Hello and welcome to MakeCast. I’m Dale Dougherty. I am joined today by the authors of a new book we published. The title of the book is AI Robots, and I have on the call Reade Richard, Brenda Shivanandan, Andy Forest, and Denzel Edwards.
Welcome to all of you today.
This is a book about how to bring AI into the classroom, how to engage kids. It really could be out of a classroom too at home or anywhere, but it’s making it practical and accessible for people that are building stuff, to just add it as another element. I wanted to do introductions.
Why don’t we start with you, Andy, and just give a brief background on yourself and then a little bit what your role in the book was.
Andy: Sure. Thanks so much, Dale. I’m executive director of STEAM Labs. I co-founded that with my wife, Marian Mader, 11 years ago now to just bring technology education to kids in the maker kind of way, just like project based and have technology serve kids to bring their ideas to life. That’s really where this book came from is just all the amazing projects that we’ve done over the years with kids through all the awesome Steam Labs team, just as culminated in this book. So we’re pretty, pretty excited about it.
Dale: Reed, introduce yourself.
Reade: Hi there. My name’s Reed. I’ve worked with Steam Labs for, God, what was it? At least five or six years, Andy? Working with them through the makerspace, and then when we went into the online version just primarily focusing on maker education, and what turned out later to be, what Steam Labs progressed into is AI education, which is a fantastic, fantastic realm to get into.
We’re at that point now where AI is becoming not just like that thing you’d see in movies, but a thing that’s actually tangible to our lives. And the fact that if we can get that into classrooms and get that into kid’s hand and teach them not only what AI is, but also like how to use AI responsibly and how to look at your programs in a very responsible way to bring people up in this now environment that it just becomes second nature to them.
That’s where I’ve been. I’m just like a maker educator out of Toronto and thrilled to have the opportunity to live my childhood dream to write a book.
Dale: Brenda, tell us a bit about yourself and what you did in the book.
Brenda: Hi, I’m Brenda. She her pronouns. I did most of all of the visual illustrations as well as photos. My background in visual design working with Steam Labs. I was able to explore new aspects of my life that I haven’t been for. So I’ve done a lot of making in the past, but now being opened and having access to electronics as well as AI I feel like that whole process is reflected in our book.
So it’s been so great working on it.
Dale: Thank you, Brenda. And Denzel tell us a little bit about what you did on the book?
Denzel: Hi, I’m Denzel. I’m a software developer based in Toronto. During my time with STEAM Labs, I did the website for the Make AI Robots for the Micro:bit section.
And it was a really cool way to integrate learning and AI into software and coding. I was really excited to work on this project. It was still the project I brag about the most, probably to this day and probably the most fun one I’ve done.
Dale: Thank you, Denzel. Andy, let’s talk a little bit about AI and education.
There’s a lot of interest, and the world sometimes divides up into pro and con in maybe unhelpful ways. It’s like debating the internet. It’s here; people will use it for lots of things, and some of those won’t be very good things, but it happens. I think there’s a lot of confusion in education like AI is going to do this or do that, and students are going to use AI to do this or that, but what I like about this book is it just fits so well with what we do at Make. How do we get our hands on the thing? How do we make something? How do we understand it by using it? Not by just creating concepts and walking through that. Does that resonate with you?
Andy: It does, and I’m glad that came through so strongly because that’s definitely our vision. We wanted to show how AI is a tool for empowering people to create stuff and just, yeah, there’s like you said that just AI is really disrupting education in lots of ways that are hurting education right now. At the same time, it’s this incredible, powerful tool as well.
The book is coming out at, I think, a really good time. It’s showing how the technology can be used. We feel that just by educating people to use it, is just the most powerful way to make sure that they can use it as a tool for themselves.
So they can talk about it. There’s all these societal questions that we’re facing now about just what is the role in AI? How does this empower people rather than replace people? To really be part of those discussions and to push for good policies by governments and companies, you need to know how it works.
You need to know how to use it. So just having this generation grow up using it and discussing, I think is a really important thing.
Dale: One of the challenges is that sometimes the education, they’re just not very familiar with, even electronics or coding or AI. People can get afraid of that; that it’s too hard or it’s too much work to understand it. What’s really great about the book is you focus it on really fun projects for the kids.
What struck me when Andy proposed the book to me is that you had a bunch of projects, but you’d actually tested them on kids. They got to do the projects, you saw that they liked them what went well what needed work, and improve them. It isn’t just coming out of your head as, this is a cool project for kids. It’s actually a cool project kids have done successfully.
Reade: Exactly. We’ve had plenty of time to prototype a bunch of ideas. I think the very simple thing which is the original title of the book was Robot Inventors, which was these cardboard robots that Andy and his team had developed. Those structures became like the basis for everything that was in the book and it was always just when we were doing kids classes at STEAM Labs or off site at other makers that we work with.
We were always testing new things and seeing how kids would use the parts to build like things. One of the things, the very first project in the book, was built just simply by kids making pieces together to throw like bats and butterflies and that’s where my mind is okay that’s simple, but it’s also teaching the concept of hinges and how to build a hinge out of like craft materials and things like that.
So that’s like. The idea with the book was using concepts like that, but getting more and more advanced as we go. We take T changes in section one or like axles and wheels in chapter two. So then these are things that like, you’ll notice they reappear later in the book.
So it’s also teaching kids like these maker skills of, okay we’ll practice something, we’ll get really good at it. And then that’ll be a tool that we can throw in our toolbox to incorporate later on in other projects that we do. And that’s always been like the modus operandi of what STEAM Labs is doing.
Dale: So each project has four parts to it, which I think is a wonderful idea. The four parts are construction, electronics, coding and AI. It’s almost like layers for that project. And you could do the project, it’s just layer one. But you could also go all the way down to layer four and engage AI. So talk about that, Reede.
Reade: I really love that concept because it’s more of a jumping off point, like if someone’s not particularly ready for electronics, you still have, if you just want to build the cardboard, you have a project right there. Then as you turn the page, you start seeing how much more fun the project can get. To use the cars example, just making a simple moving car out of cardboard with wheels, that’s a great project, but then you see it’s oh wait, I can turn this into even more of a car by adding just two simple LEDs to make headlights.
And then from there, it’s oh, kids are learning Micro:bit in school right now and Scratch programming. So the idea of make code just I’m finding with a lot of kids that I work with even today, they find like Blockly is just a perfect language for them to put things together.
So that’s why we decided to incorporate like Micro:bit because it’s just we’re on a process of automation. Like at that point, once you get the electronics, okay let’s find a way to control the electronics so we don’t have to do it ourselves. But then Andy came in with this great concept of, hey, we found a way to control the Micro:bit with AI. Let’s do something with this. Let’s now automate the Micro:bit process by just allowing AI to control the rest of our project. As I said, it’s kinda like the book is all about like ramping up your skills, but even chapter to chapter, it’s like you’re starting to test yourself by oh, let’s just build more of a simple cardboard project.
And then as you go through. Adding electronics is like the next step in a maker’s journey, and then how can we further this? Let’s learn to code, let’s learn to take this to the next level.
Brenda: I think what’s also really powerful and what we spoke about earlier is how that progression also reflects how we are as children learning as well as parents who there could be immigrant parents who don’t know anything about electronics or about AI let alone.
I think from this particular book, it’s how figuring out the relationship the student has with their parent, and that moment where the parent the child, by learning about Micro:bits, by learning electronics, and about AI, there’s a teaching opportunity that isn’t being seen quite often where the child is actually teaching the parents and then also how they’re learning together simultaneously. We’re embracing that relationship as well.
Dale: Explain the AI Micro:bit connection here, how that actually works in the book. So you’re, it’s, if I understand it we’re not putting an AI library on a Micro:bit. We’re actually accessing that through a cloud service, right?
Denzel: It’s really cool because the way it’s integrated with Micro:bit is in a way that really allows kids to understand what they’re doing while they’re doing the project. Like the project where you wave and suddenly the sensor would go off and I showed it to my little brother once and he was like, whoa, like how through a camera, how did it recognize that?
And how’s this little computer understanding? And then spent the next hour I spent explaining it to him and he was just enthralled. I was like, Oh wow, this really works with kids and it really makes them interested. It was really cool because. I could explain it to him it wasn’t a hard process.
It was a step by step process that uses components that are in front of them, the Micro:bit, and then cloud services that allow it to easily integrate with these different projects and do these really cool things.
Andy: One thing I might add too, just around on the way it works. We’re using a vision recognition system that runs completely in your browser with your webcam.
So you don’t need anything more other than that for doing the AI part. And then he’s built this tool that bridges that. vision recognition system and just sends it right into the Micro:bit. So teachers are already teaching Micro:bit in class. There’s millions of them out there in the world. So now this tool that is already available now gets this extra sense, get this extra input of being able to see the world and hear the world and make decisions inside of their code based on that. It’s integrated to the level where you train like Denzel was saying, like putting your hands up in the air. The kids can just teach it to recognize that by taking pictures of themselves doing that. Just press the button with their hands up in the air and they train it to recognize that.
And then it’s just fed right into their Micro:bit. With an if statement that they’re already familiar with, if they know Micro:bit, or if they’ve done step three of the four steps in the book. They already know how to do if statements in Micro:bit, and they can just put in if hands up, then do this with the Micro:bit.
So it’s a really approachable way and we’ve just developed that kind of learning pathway over the years. Like Brenda was saying, it draws them a little deeper. Here’s another capability, you just have to learn this a little bit more, and now your project can do this extra thing.
Dale: It’s so different than just, say, maybe typing in a prompt and getting a result back. It’s really an interaction, like you were talking about the hand waving, computers recognizing something that you’ve programmed it to know or learn and now it’s acting on that. You get to see how well it acts and sometimes you have to improve it. Interaction is a pretty big component.
Andy: Teaches systems thinking too, because that;s where the AIs are being put to work in the world. The ChatGPT, raw by itself, doing your homework. That’s not a good use. That’s not an ethical use, but integrating these different things together into a system that can accomplish your goal in an ethical way, that’s what we want to teach that’s the future.
Dale: Reed, did you have trouble deciding on which projects to put in the book and which ones work best and which ones built up to greater levels of complexity.
Reade: To be honest, I have enough content for probably three or four more books. That’s what it came down to. Over the time with many people looking at it was like flow and then looking at which projects could be done with more basic and intermediary steps like my favorite to be honest is chapter 3’s ball toss game. That’s simple but like to build and then that’s where the code starts getting into more difficult stuff like doing timed loops and things like that to start challenging your code. But you know getting up to something that’s your chapter 7 wearables where you’re starting to create like superhero wrist stuff or target practice games. These grandiose ideas that I had that kids have created parts of these throughout classes throughout the years that I’m just like all these are awesome but they’re too difficult to start off with. So it was more of a balance of figuring out like which projects can be done at which varying skill levels. I know, say the superhero shield has servo motor movement in it, I know it needs coding and I know it needs like a certain build mechanisms.
Okay, so what are those? And they created like a chart for myself being like, okay this concept can be brought into this chapter so that I can rely on it later on. Like that whole path of using what’s already in your toolbox. So by the time I get to chapter 7, like if readers can realize I become less hands on with the code and kind of be like, you can figure this out. We’ve already done this. If you need help, the photos are there, but by this point, if you’ve been following along, you have the skills. That’s like the part of the book that I’m really proud about, especially what Brenda talked about, about like kids being able to teach their parents. I feel whenever kids show these things to like the adult generation, I was like, I don’t understand this, but I know my kid loves Micro:bit and now this is a great opportunity for, it’s okay if I can work with a kid and now we can, the kid can become the teacher and have this whole new level of kind of like responsibility and feeling about them to feel empowered about the knowledge that they have.
Dale: Andy, did you want to say something?
Andy: I was just gonna say just about the empowered step like we tested a lot of these on my son, who’s nine now and so he while working on the book for about four years, I guess though. He’s grown up with some of the projects. But I love now like he has this constant battle with his older sister who, his older sister puts this picture upside down in the kitchen without anybody noticing, and then he son Rory puts it right side up again, and it dawned on him one day, after he’d been learning these tools, that he could actually use the things he learned to make an alarm for when she did that.
So he programmed the vision recognition tool to just recognize when that picture was turned upside down and just notify him. So he’s always just seeing the world in a different way now. There’s challenges and then he knows another kind of tool in his toolbox that he can approach those challenges.
Dale: One of the great ways of looking at projects is their patterns for designing something, templates, whatever word you want to use, you understand how that works and you come up with a different idea to apply that pattern, right? Oh the camera recognized my hands. Can it recognize something else? Now I know how to make that work. And I almost would wish teachers that use this to use that time, like once you’ve built a basic project, ask them to riff on it, ask them to improvise and come up with something that’s uniquely their own.
Reade: I mean you can actually see that within the chapters themselves where we had that idea. I grew up with old Lego sets and in the back of Lego sets they would show you like, here’s another thing you could make with all these pieces. So that’s why we’ve incorporated the taking it further section to look, I’m a maker, I’ve built like these little devices using the concepts and the same coding you’ve done in this chapter, can you do the same?
Like I, that was like a whole impetus, just to reinforce kids like, We’re teaching you something here, but it’s useful in other ways. What can you do with it?
Dale: That’s when they begin driving the process a lot more. And that’s very exciting.
Andy: I love what Reade’s done with them too, is there’s always in the step one, the build phase, there’s always some creativity, like even right from chapter 1, he mentioned butterflies and bats, but what he actually designed with Brenda is actually a hinged creature mechanism that doesn’t have the wings, any of the character on it on the kind of core mechanical piece. So you have to add your creativity. What kind of winged thing is it? And so the, they’re all designed that way too.
Dale: Let me ask you, there’s a challenge with AI, writing about it, putting it in a book, as it’s changing so fast. How did you think about that going into something that takes, really a year or more to put together?
Andy: That’s a good question. So we wanted something useful right now, obviously. So we wanted testable things, but the capabilities are advancing so fast. Since the tools are digital for some of us, is we do actually have a plan to improve some of these as we go along. We’re building lots of other AI integrated science center exhibits and websites and other innovations. We’re on the leading edge of all those tech things, so we actually have a bit of a development roadmap for improving the online tool to bring more AI things into it.
I won’t give too much away from that because I don’t want to over promise. We definitely want to be bringing more AI capabilities into that tool so people can continue to explore the latest AI tech for their projects.
Dale: That’s great. But I hope people recognize the value of the hands-on pieces here, doing these projects that have characters or animals or other things. They really capture kids’ imagination. And they get their attention, and then you’re adding these other layers later on, and like Denzel was talking about, your brother or something is fascinated by how this works.
Brenda: There’s four sections for each chapter, three of which aren’t even AI. It’s more so like the building part, learning how mechanics works, and then we get to the AI. And I feel like just how the book is structured, it’s more so embracing the learning steps and the learning opportunity between all those stages.
Yeah. And even if you’re like, when it does advance, you at least you have the foundations so that you’re well prepared to take the book further and the projects further.
Dale: For all the use of the word STEM and STEAM in the world so often these things are treated as different subjects. In this book, they’re really integrated very well. They’re learning mechanical things. They’re learning electronic things. They’re learning coding things and now learning AI things. I think it’s very unusual to teach in an interdisciplinary way and to create projects that cross boundaries.
Reade: Part of the research for that as well was like not only going through like the backlog of what have we done at Steam Labs, what can we riff on, but I also talked to a lot of my teacher friends, and a lot of the teachers that we worked with. What is something that like you could really use in the classroom, what are some subject, what are some concepts that you think would be great that isn’t really being covered in the curriculum as much as you want. Four out of five teachers would always tell me linkages, so that’s why the whole idea of the book started with using linkages to show change in motion. Then from that everything spitballed and became larger than life from there, but like I said like reaching out and finding some of these things that aren’t being taught, and how can I help assist these teachers or people who could use this material in a way to go from there?
Dale: All of you have worked a lot in the, what’s sometimes called the informal learning sector, science centers and after school and all of that. You do have a greater degree of freedom sometimes to do things. And, but what often seems in a really good way, what matters most is, are the kids interested?
Can you get them to do this? Whereas school sometimes says they’re not interested, but we have to do it anyway. I just believe teachers can learn a lot from informal ed by the things that work there and bringing them into the classroom and maybe you can hang other things that you want to off it. But the engagement is one of the key things.
Andy: I think you’re absolutely right. And that’s one of the things I’m most excited about for AI support for education is because integrated learning is really hard to do to bring all these different things together into a project that has different curriculum subjects and different technology and technology that a teacher might not be familiar with.
One thing that ChatGPT is actually pretty good at is helping with bringing those things together. So we’ve done some experiments with just having a teacher say, ” I want to do a project- based activity. I need to hit these kinds of areas of the curriculum. I’ve got this book that shows me how to make Micro:bit projects. Can you give me some ideas of how I could bring that together with maybe the space science unit that I’m trying to teach right now?” And it’s pretty good at coming up with some ideas. It’s also good at helping teachers to troubleshoot and adapt stuff. Like the Micro:bit coding section, it has a text based version of the block coding that’s all in JavaScript, and you can take that and copy and paste it into ChatGPT and say, “Hey, I want to do a different thing, but I don’t know how can you help me?”
That’s one of the. things I’m most excited about is just these AI kind of helpers for teachers on how to bring integrated learning together in a project based way and then also fill in some of the gaps in their knowledge to help them feel more comfortable to bring it in front of their classroom and then support the kids projects as they’re going.
That’s another gap that teachers frequently have. ” The kids made this Micro:bit project and then they hit a bug that I don’t know how to help fix” and ChatGPT actually can, which is, I think very empowering.
Dale: There’s something unique to AI, which we mentioned already, but I think it’s this idea of training. It’s just a fascinating thing.
When you’re just doing coding, you’re just giving instructions or commands to a computer and you expect it to just follow those commands and do what it’s expected to do. Training is almost a little bit like training a human, right? You don’t know what they understand and you don’t know exactly what they’ll act on. So you have to perfect it, iteratively. So I think that’s a really new thing for kids to work with here.
Reade: I love that concept. And that’s the challenge of trying to teach AI while not being physically in front of someone; you can’t really explain why their project didn’t work.
So part of the process of going through AI in each chapter was explaining, okay if it didn’t work, try to take more photos, try to make sure your photos are diverse. So It’s imparting these tips. In the book if your training model doesn’t act the same ways that ours do, we also give these tips. Each chapter like new tips about okay, how can we improve upon this? And also specifically, why did our model fail? It’s because we did X, Y, Z, but we didn’t think about A. I like the concept of letting someone fail to learn a lesson specifically. As much as, kids don’t like failure, they just love success.
But there’s the learning and opportunity that comes with, okay, this went wrong because we didn’t think about this part. So it just reminds them at that point. So the next time they’re doing it, they’re like, oh, wait, so maybe I got to look at this from another angle. And that kind of becomes like not only a tool for working with AI, but a tool they can use like lifelong..
Dale: Andy, you mentioned your nine year old son doing the projects in the book. This book looks like something that kids could go through somewhat on their own with some assistance. Partly Brenda, the way that you made it very visual. This is not dense technical verbiage. It’s pretty clear. You can see what you’re doing along the way. It’s very explicit in the steps. And that helps someone who’s maybe unsure if they’re able to do this to just do it.
Andy: Yeah, we wanted to make it really approachable and and usable from a young age and also interesting and usable from an older age too. So I think it does serve a really nice wide age range. And of course, when you’re getting into the coding and then we usually aim for kids that are perhaps eight and up, but younger kids with some assistance can also do even the coding steps.
We’ve worked with kids as young as grade one, like six years old doing vision recognition kind of projects with, and they told us in, after workshops that they enjoy learning to think like the computer does and and I think the younger they can be exposed to that kind of computational thinking the better. They know how to frame their problems in a way that can be solved by technology, as they’re growing up.
Dale: That’s really fascinating. The materials that you’re using in the book are pretty much cardboard and simple things. Micro:bit is a common microprocessor, very cheap. It’s all over education. It’s very affordable in a way to be able to do this with the individual kids, not just like one class doing one project, but individuals doing them in that class. Sometimes I see people starting makerspaces and they want to know what tools to get. And that’s not the biggest problem they’re gonna face is like how to get the know how around those tools to do things, how do you even put a car together from cardboard, and once you see it though, you do get it, right? That’s where I think books are really great at that.
Reade: Yeah I, honestly the materials — this is what Steamlabs has always worked with. It’s just like these easy to find, as much as plastic straws are going away, like having those pieces there or wooden dowels or pencils or cardboard or mostly recycled materials, like beyond the electronics, which I wanted to make sure were all like easily accessible. It was like, at the very sourced on Amazon for cheap. Or if you have a local electronics store, all the better. Even then I didn’t want the projects to have parents to go out and okay, I have to buy all this extra materials to build the things. They should all be things that you have around house and especially, as it was being written during the pandemic, we all had tons of cardboard from all the things coming in every day.
So that was the impetus is just, we have all this material. Let’s not put in the recycling bin, let’s recycle it ourselves into projects.
Dale: Thank you all for talking to me today about your new book. I’m very excited to publish it and to see it out in the world. It is a it’s a great looking book. I really hope parents get into it. Sometimes you don’t get this stuff in school and your kids should pick this up.
So thank you all any last words from any of you?
Reade: I just want to say thanks for the opportunity for publishing our book and helping us get these ideas out there to the public.
Thank you for listening to Make: cast.
ADVERTISEMENT
