AI Explorations and Their Practical Use in School Environments is an ISTE initiative funded by General Motors. Through professional learning opportunities for educators, the program is designed to prepare today’s students for tomorrow’s AI careers.
Recently, we spoke with three participants of the AI Explorations program to learn about its impact in K-12 classrooms. These innovative educators discussed the challenges of completing the program and their experiences planning and implementing AI curriculum in their classes.
Coral Zayas is the new elementary instructional support specialist for science in Crowley ISD’s iNetwork of schools. She was an ISTE-GM AI Explorations program year 2 participant and has since implemented projects from the Hands-On AI Projects for the Classroom guides in her bilingual classes in Leander, Texas.
ISTE: What motivated you to join the AI Explorations PD program?
Zayas: I knew that it was a perfect chance to learn from other educators and experts in the field. All of the ISTE groups have phenomenal educators. So, I was motivated to take the industry and peer knowledge and then bring it into my classroom.
We tend to call AI emerging tech in K-12 or higher ed. But I know that these technologies have existed outside of the education space much longer. And I get to bring that industry information to my students and help them make those real-world connections.
That’s why we did an AI unit last year—to show students this new and exploding field of technology and computer science in class and help them see those connections. The younger I can get them interested in emerging fields like artificial intelligence, the more exposure they can have. Then, they can develop their interest in taking other STEM courses as they continue in middle school and high school. I think it’s meaningful to bring what we see outside of education into the classroom.
When did you start implementing your own AI projects based on what you learned in the AI Explorations program?
I got to teach a brand new course in our district last school year. We were piloting and exploring a STEM course for sixth grade students. We had a computer science unit that was built into the course. So, I told my sixth grade teammates about my interest in artificial intelligence and asked, “Do you mind if I add a mini-unit that we can attach to the computer science course, which focuses on artificial intelligence?” They were very excited about it, and they said, “Yes! Let’s try it!”
Using a variety of resources, I created about four weeks of AI lessons with different activities. My class had an interesting discussion based on “How I’m Fighting Bias in Algorithms.” The ISTE-GM AI Explorations network helped connect me to other resources outside of the course, too. One of the things that the Facebook group shared was MIT’s new AI and CS community for educators. I was able to get access to many lessons on artificial intelligence and implement those lessons in the mini-unit as well. I used quite a few of MIT resources like Dancing with AI and Zhuori, as well as ISTE resources.
What were the ISTE resources that you included?
We used some of the lessons from the free elective teachers’ guide and the guide for secondary teachers. I also combined some resources to match what we were working on at school, for example, UN sustainable development goals, which we worked on in other projects in the STEM course. We combined those lessons to reiterate other learning experiences, connecting them to artificial intelligence.
Amanda Bailey is an African-American team leader of the ISTE-GM AI Explorations program and the district technology coordinator for Crescent Academy in Southfield, MI. Crescent Academy is a Title 1 school that serves 90-99 percent African-American students. Amanda’s team developed a capstone project for elementary students: AI and Machine Learning for 5th Graders.
ISTE: You started the AI Explorations program during the pandemic. Can you discuss some of the challenges you experienced? How did you get through them?
Bailey: At first, looking at what was expected of us, I thought, “I can’t finish this.” And then I said, “No, you’re not going to give up, you can do this, you’re working from home…” So, I just made the time, even if it was 30 minutes here and there.
I set a virtual meeting schedule with my teammates, Matthew Blacker and Rasahn McCombs. And we all collaborated via Google Docs or Slides before and during the meetings. Then, we came back together and met before the due dates. Once we did it the first week and got the hang of everything, it was a lot better.
My team supported me a lot. Having that support of the whole group on the ISTE platform, connecting with other groups and team leaders, and especially having Steven Jones as our coach helped me a lot. Steven was really supportive and made himself available whenever we had questions.
Now that you’ve completed the AI Explorations PD program, will you be implementing an AI learning project?
I have plans for a maker space that is going to launch in January. I want to incorporate tools and resources from Google Earth, Tinkercad, Code.org, ScratchJr and others to be used with Chromebooks and iPads. This space will be for the Pre-K through first grade levels.
I am sorting through the tools that I learned about from the ISTE-GM AI Explorations program and trying to figure out how to help teachers use them with their students. I would like this to be inquiry-based design and based on challenges as the instructional guideline—using tools and resources I find in those AI Explorations course modules that fit the guideline.
Would you please share your experience supporting a district with a large proportion of students of color?
In the past, materials were limited and supplies were limited. Teachers didn’t have sufficient knowledge. And, just as I was getting ready to train my teachers, the pandemic happened. So now, I hope to use what we have available to help the teachers in my district support our students.
When I was getting my degree in educational technology, I learned that you always consider what you have first and what you can do to modify it. Not every school has a Mac or iPad for every student. Using what you have to make it work is important. I am able to do that with the content of the AI Explorations course. I am excited to help teachers teach coding while telling a story, thinking about how we can intertwine literacy and STEM. For example, ScratchJr is one thing I want to do in the maker space. We don’t have to build a robot. Let’s just use what we have available to get students to think, tinker, play and test their ideas. Play is the highest form of research. I want to take pieces of hands-on experience from the course and apply those STEM concepts.
Eamon Marchant is a forward-thinking high school teacher, coach, site tech coordinator and science department chair at Gretchen Whitney High School in California. He was an ISTE-GM AI Explorations program year 2 participant and has since been teaching about AI topics in his AP CSP class. His site also offers an advanced high school course focused on developing and programming AI applications.
ISTE: What sort of challenges have you met when implementing AI lessons and projects with your students?
Marchant: There are a couple of difficult little things. One of them is that many AI lessons are accessible, but they are very distant from the actual code itself—far away from the actual work it would take to make something. The kids pick up on that. When I show kids something simple—for example, the Machine Learning for Kids site and Google tools like Teachable Machine—they understand it, and they like the ideas behind it. But, there’s a lack of satisfaction when they can’t build something themselves.
To find ways to make students think, not only “Can I use this tool?” but “I can build tools myself,” there’s still a deficit of resources. And it’s not anyone’s fault. It’s just a challenging jump with all the math and code that goes behind it. And I am still trying to figure out ways to bridge that gap.
How do you help students face those challenges?
Our strategy is a two-pronged approach. We use Machine Learning for Kids to just have them get comfortable with the concepts first. Then, when they go on to our AI-focused class, they start from the absolute fundamentals of programming. So, they begin with organizing lists and learning Python syntax. From there, they can work their way out to writing a neural net.
We want to see if they can connect the dots when they get to the end of the course. Hopefully, I’ll be able to say that they put together all the things they learned and understand it this year, but that’s conceptual.
Where do you most often see students struggle when learning AI?
I think the biggest hurdle for many students is that, when they first dive into using AI as a tool, especially from the code approach, a lot of them aren’t used to a computer not doing what they want it to. Most computer programs they’ve been using are either super intuitive or super forgiving. But writing code isn’t like that. So, when they have an idea written out, but it doesn’t work, it’s a whole new type of frustration for many of them.
This is something new that they have to face. If this was a math class, and they got a problem wrong, that’s okay. They can still turn in their homework the next day with that wrong solution. They might not even know that it’s wrong. However, this is a different case. If the program doesn’t run, it just doesn’t run. Often, students feel like they can’t turn in something that doesn’t run. I have to convince them that they can, knowing that I will at least give some credit for their effort.
Any AI learning resources you’d like to recommend to educators?
Some of my personal favorites are Experiments with Google, Blob Opera and Magenta.