Technology Essentials in Education Episode 35:
What Happened When Kindergartners Learned About Machine Learning

Welcome to Technology Essentials in Education. Today, I'm chatting with Jessica Holloway, who spends her day side-by-side with teachers. She's an instructional coach, and she has a lesson that she co-created with a kindergarten teacher using Google Teachable Machine, and they've taken this lesson, their work from their own building, made it all the way to the national round of the Presidential AI Challenge.

So, we're going to get a behind-the-scenes look at this lesson, what it looks like, and how you might use it with kindergartners or really any elementary students or students who are exploring AI for the first time. Let's get into the conversation.

This episode is brought to you by Jotform. Jotform provides an all-in-one solution to streamline administrative tasks, enhance community engagement, and foster innovation. Using their no-code drag-and-drop Jotform and workflows, your teams can securely collect and store data, automate tasks, and collaborate on team resources. Educational institutions are also eligible for a 30% discount on Jotform Enterprise. Head to their website to learn more at Jotform.com slash enterprise slash education.

Welcome to the podcast, Jessica. I am so excited to talk to you. Jessica Holloway talks to you about computer science and AI, but specifically on how this all connects to early learners and what we can kind of take back into our practice no matter what groups of students that we work with.

So before we get into all of that, can you share a bit with listeners? What is your role in education? What is your day-to-day look like?

Hello, everyone. I'm Jessica Holloway. I'm an instructional coach, and on the day-to-day, I spend a lot of time collaborating with teachers, co-creating lessons, co-teaching, coaching. I like to think of myself as an idea pollinator, so I take good ideas and share them around, but a lot of it is side-by-side work with the teachers in the building.

I love that framing because I think it really is so much of what coaching can be. We have an idea. We're bringing it to someone. We've heard them talk about a problem of practice or something that they're trying to solve in their environment, and we bring these, hopefully, solutions to that conversation. So I think that's such a great illustration of what that work can look like.

And today, we're diving into a topic that I think will pique a lot of listeners' curiosities regardless of the age group that they work with. So if they're kind of working with young learners now, they're working with middle school or high school students, everyone's kind of impacted by what happens in early learning environments.

So as we think about AI and computer science with really young learners, can you tell us about the AI lesson you've been doing with kindergarten students and what inspired you to bring this topic to that age group?

Yeah, usually the first reaction when I tell people, yeah, I started out doing AI lessons with kindergarten, they just kind of pause and have a stunned or confused look like, are you sure that was a good idea?

You know, this really started when I had an opportunity to participate in an ISTE AI Explorations course. It was when I was just figuring out what is AI in general. I had a real hesitation of like, I don't, I can't make AI. I don't really understand AI, and working through the course was really helpful. It was a team of us and it was, we had representatives from K-12 and coaches as well. And so we could have that vertical conversation. What might this look like from kindergarten to our seniors? And so it was a really great experience. We were very fortunate to be able to participate in that.

But one of the assignments focused on machine learning and Stephanie Elgin, the kindergarten teacher, she actually works K-5. We had just been doing different lessons, but they just finished up learning about living and non-living in kindergarten science. And we had just learned how to play with Google Teachable Machine. And so we thought, huh, I wonder if we could do something with this with our kindergarten students, because it fits so well with classification of data.

So from that idea sparked a lesson that we developed and said, all right, we believe that students are expected to understand the characteristics of things that are living and non-living. So our lesson, we start with a data sort. They sort into the two categories. We review the characteristics of each. We do some self-checking and then we do have an introductory video. It's very short about AI just to get the language for them and get an age-appropriate introduction.

It's a cartoon. It talks about how students learn in school versus how robots and machines learn. That was really important for Stephanie and me in making sure that we're distinguishing how humans learn versus how machines and computers and robots learn. And so we give them that language. Then we introduce Google Teachable Machine and do some front loading of making sure the devices are ready. But the students have their data. Their data is ready, and then they input it.

They are scanning the pictures in. They have a partner, one of them's holding, one of them's pressing, they swap and they swap categories, they get all the data in. But we did iterations of this. We did have to start having conversations about what's clean and dirty data because in our living, it's okay if your face is kind of in there, but you can't have your face in any of the non-living data because then it's going to confuse the machine.

So that actually unexpected Teachable moment came up. We did get to have conversations of like, go ahead and get your selfies out now in the living section. Do the selfies. We know you're the one to do them, but you can't have it in there. So that was a good conversation. Then we talked about now they trained their model, then we have to test it.

We come back together as a class. We test it out. We talk about, is it right? Is it wrong? Why are we not getting the results? Do we have enough data in there? So it's a really great conversation and introduction to them about how machines learn and they have to be given the data versus how we learn when we go to school with experience and all of that.

So I'm a firm believer those kiddos can say Tyrannosaurus Rex. They can say AI and have that conversation with them. And that will set them up to continue to learn and talk about AI as they move forward.

Wow. What a powerful activity, and of course, that T-Rex example, too, right? I think humor is at home, but what a powerful activity, because you can even just, you know, as you're talking, my wheels are spinning about what this would look like in a second-grade classroom, what this would look like in a fourth-grade or an eighth-grade classroom when it comes to maybe finding those data sets themselves, or talking about these things with a partner, or even a sixth-grader having this conversation with a second-grader. You can almost see them being like that buddy taking them through.

So just so many ways to take something that absolutely feels intimidating on the surface and bring it to a group of students who probably are more willing to try and fail and then kind of the things that maybe a middle schooler would feel less vulnerable jumping into and trying out.

Yeah, and I would even say the teachers are less comfortable jumping in and trying. And so there is a part of the work with that that Stephanie and I have been trying to do. We've taken this lesson to several conferences, sharing it with teachers, but first letting teachers experience it as a student, building their competence and confidence in what AI is, what machine learning is, before they can turn around and use that with students.

And we entered this in the Presidential AI Challenge, and we won the state level, we won region, and we are going to the national.

We are excited to share this with more people. We are excited to see it's resonating a lot about how we can demystify AI for our younger learners.

What a fantastic journey from something that might have felt kind of small or in its own little room with 20 or 30 students to something that's been able to reach so many different classrooms through just the work and the storytelling you're doing around this lesson.

When we think about the connection between AI and computer science at this age, are you weaving some of those foundational computer science concepts like patterns or sequencing alongside this AI piece, or are you keeping it separate?

Integrated all so too often we try to isolate so much in education when so much of it is really integrated. And why not let it be that way? And honestly, K2 is the perfect grade and age for computational thinking, which is really about how to get kids to think and that thinking will evolve and become more sophisticated as you get into computer science, AI, but all content concepts.

So in computational thinking, we have that pattern recognition, abstraction, decomposition, and algorithmic thinking. So all of those pieces are what they're doing right now in K-2. So pattern recognition in certain types of words, you know, do you have CVC words? Pattern recognition right there in learning in literacy and pattern recognition in storytelling beginning, middle, end. And what information do we need to know to comprehend and not need to know? How do we take this math problem and break it down into smaller pieces? Like all of that is naturally happening in K-2 if we just put the language and notice and name it.

And that idea of being intentional, the notice and name it, using that vocab with students who are like sponges taking in all this information in these different contexts. I love that connection between two things, especially two things that someone might not feel like they have the capacity or the confidence with. But now they're diving in.

And when you're designing a lesson like the one you described for such young learners, what is the decision-making process when you're trying to figure out what's developmentally appropriate? Are there particular frameworks or approaches that you lean on to make sure that the content is truly accessible for that age group?

Well, I lean on teachers because they are the experts, and it's interesting sometimes like we look so often outside of buildings for experts, but the teachers that are in the rooms for those students every single day, the expertise that's in there, I try to tap into that. So if I'm building a kindergarten lesson, I want the voice of those kindergarten teachers in there looking at it, but also considering what are our grade level standards. Also thinking about what's developmentally appropriate as far as thinking skills, physical skills as well, because some things like they are still developing hand-eye coordination. So some skills may be a little trickier to do at certain ages, depending on mental and physical development.

So just keeping that in mind and then also just trying it out. And then you may realize, okay, this may not be best for first grade. This may be better fit with third grade. So sometimes it does end up being a trial and error kind of situation. But that's why it's really great to have teachers who are willing to try things and give you the feedback and kind of take that innovation journey with you and say, yeah, let's see what our kids can do.

But I would always say don't underestimate those K-2 students. Yes, we are learning how to read that. They don't necessarily have the independent reading down, but they can do so much thinking and they are grasping vocabulary. They're getting those literacy skills. So don't be afraid to tackle big concepts for them.

And just like you mentioned about leaning on the K-2 teachers who are ready to give feedback and let you know what's happening in their classrooms. We know that students are often happy to give feedback and share their feelings, particularly at the elementary level. And maybe they might turn a bit more inward in the second half of their K-12 journey.

What kinds of reactions have you seen from students during or after this lesson? Have there been any moments that have surprised you or really showed you that these young learners were making meaningful connections to the concepts?

Yeah, I think the excitement of when I watched the kindergartners interact in that lesson, they took on the responsibility of being good data input people. I was just surprised at how invested they were in the process, and they wanted to know, is it going to work? Is it going to give us the right answer? And so just think, surprised that they can zoom in and focus because they are squirrely at that age. You know, they got a lot of energy, but they took it on and they were excited because sometimes they don't get to work with computers as much as our older students do.

They don't necessarily have them in hand. And we were using Chromebooks and a lot of times in the younger age, it's iPads or tablets. So new hand placement, new things, new learning. But with teachers and coaches and principals, they all just kind of have this positive, like, really, really? We can do that with kindergarten. And so I think it's almost just like breaking this misunderstanding or myth that they can't. They're too young. We can't talk about computer science. We can't talk about AI.

But I think that's just a myth, and you do have to be developmentally appropriate, but they can understand the concepts. We can approach it in ways that are accessible and not intimidating, and it doesn't necessarily, you don't have to have a device. You can do things that are unplugged, analog versions of things, and then you can introduce technology and tools as needed.

Yeah, and just that thoughtfulness of why we're bringing this to a group and what that's going to look like. And I really like what you said about the students' investment in the process. The same way that they might get excited about something happening with a character in a read-aloud. Like, they're going to be upset if it's time to go to recess and you don't get to finish that story. Anyone who's done a read-aloud with that group knows they want to know what happens at the end.

So, I think that's such a great comparison to, like, is this going to work? Is this system that I'm training, that I'm interacting with? So, I love that kind of buy-in from the group.

So, you know, if there's a coach or specialist who wants to bring this lesson into a classroom that isn't their students, I'm using air quotes if anyone's listening, like their kids or their class, what's your advice for collaborating with a classroom teacher to make this a smooth experience?

Yeah, I think anytime you can be right there alongside a teacher through the entirety of the process, it's gonna make them feel safe and you're gonna get to see firsthand how it unfolds. And you know, there's always something about seeing it for yourself versus someone's version of it. And so when they're teaching, you can scan the whole room, you can pick up on things that are happening. You know, you're right there with them.

So I would say try to go through the whole process with them. Don't be afraid to take risks and know that sometimes it's not gonna work. Like sometimes you're gonna design a lesson and five minutes in, you're gonna go mm-mm, this is not working, but that's a learning process. Learn from that and figure out what went wrong and then come back and make it better.

And I think it's definitely an ongoing process and it's one that I'm continuing on. I'm getting to participate in a community practice in ISTE for STEM in K2. And we're tackling that question of like, what does STEM, like good STEM learning look like in those early years? And I am so grateful that there's a group having this conversation because it is helping me think through what's already out there, what problem do I want to tackle? How can I zoom in on a piece of that problem?

But find your people who are willing to have the conversations, try things out, get you feedback. So create, even if you don't have a formal community practice, like create your own and then work through it and support each other as you work through these innovative ways. See what the students can do.

Yeah, just having some collaborators right along the way is huge. And for a listener who's excited about this idea, who's listening into our conversation and wants to bring this back to their learning environment, what's one small step they could take this week or this month to begin exploring AI or computer science concepts with their students?

I would say, especially if you're looking at the younger students, look for opportunities to bring in computational thinking, because that is naturally happening everywhere. I would say it's naturally happening in K-12. So you could pull that language in anywhere. But then think about opportunities to connect concepts from whatever content area you're in with concepts you see in computer science or AI.

Because, again, we like to isolate things like, oh, computer science is over here. My English class is over here, but there's opportunities to connect the two and really make that learning and thinking very sticky for students. So look for opportunities to integrate, and if you're just not feeling confident in your own understanding, look for opportunities to grow in your own. There's lots of free, completely free tools out there to better understand AI and computer science.

So look for those opportunities and give yourself permission to be a learner and a beginner in this process. And I think that is something that's a good reminder for anyone who especially is trying out something new or something that might feel a little intimidating for the first go of it, particularly when you have your group of students in front of you and you want things to feel ready for them.

So such great advice, Jessica. And for anyone who is listening, wants to continue following along with the work that you do, that wants to learn more about this lesson and resources kind of around this topic of supporting students this way, where can people connect with you? Where can they learn more about your work?

Well, you can find me on LinkedIn, Jessica Holloway. And then most of my social media handle tag is Holloway Reader. I'm a big reading fan. And so when I first made that tag, that's what we went with. So that's where we're going. Most platforms have the same handle on there, so I look forward to connecting with everyone and continuing this conversation. And if you don't have a group of people, I'll be happy to be part of your group of people.

Well, thank you so much for your time, for just all of these great ideas for someone who wants to bring AI and computer science to their learners. It was lots of fun to chat with Jessica today. Let's go ahead and finish up like we always do with a few takeaways from the conversation.

First, integrate computational thinking into content areas you're already teaching instead of treating computer science as a separate subject. Lean on classroom teachers as the experts they are when designing lessons for age groups outside your own experience. Use the language of AI and computer science with young students since they can handle big vocabulary like T-Rex when it's introduced in context.

Walk through the entire lesson process alongside classroom teachers to build their confidence with new concepts. Thanks again for listening to this episode of the podcast. Make sure to connect with Jessica if this is a topic you'd like to explore further. A big thank you to Jotform, the presenter. To learn more about Jotform and how educational institutions can get a 30% discount on Jotform Enterprise, head to Jotform.com slash enterprise slash education.