Note: This is part of a 7-video series with Dr. Kecia Ray in conversation with industry experts on using artificial intelligence in the classroom.
Welcome back to AI Insights: Conversations with Dr. Kecia Ray where we connect with experts in the educational technology field to stay on top of trends in artifical intelligence (AI).
On today's episode, Dr. Kecia Ray is joined by Joseph South, a national educational technology leader who serves as the Chief Learning Officer at the International Society for Technology in Education (ISTE). He is the former director the U.S. Department of Education's Office of Educational Technology, focusing on evidence-based learning transformation and the advancement of digital learning initiatives.
A conversation about the global role of AI with Dr. Kecia Ray and Joseph South
A full transcript of the episode appears below; it has been edited for clarity.
Discover best practices for integrating AI in the classroom.
Kecia: Hello, everybody. Thank you so much for joining us today. I’m Dr. Kecia Ray, and I am here with the infamous Joseph South, and he is going to give a little introduction to himself. I know probably everybody listening knows who he is, but just in case you don’t, he’s going to share a little bit about his background. He’s at ISTE now, but I would like for you, Joseph, to share your background and all of the amazing experiences that you had before you got to ISTE.
After that, we’re going to dive into a conversation around AI. So, take it away.
Joseph: Thank you for having me, Kecia. I’m so excited to be on your podcast and to be part of this experience. I am currently the Chief Innovation Officer at ISTE. Previously, I was trained as an instructional designer and I’ve developed all kinds of instructional courses and materials, everything from pre-K through higher ed. I even spent some years developing adaptive language learning systems, which was a really fun thing to do because it’s a hard problem to solve.
I had the incredible fortune to work in the Obama administration in the Office of Educational Technology and was the director of that office at the end of the administration. Part of what we did there was create and release the National Educational Technology Plan.
That’s an EdTech strategy plan for the entire country. We also did a lot to promote open resources in K–12, and we did a lot of work to make sure every school had broadband in the classroom. When I left the Department, I spent an incredible year at IDEO, a very famous human-centered design firm where I got to work on a lot of super-interesting projects.
The one I loved the most was called the Purpose Project where I got to help young people identify and articulate their purpose, which is something that matters deeply to them and to the world around them. I found my way back to ISTE—I say back because Richard Collada, who I worked with in the Department, became the CEO of ISTE.
And through some wonderful friends like Kecia Ray, who was on the board at the time, of ISTE, I was able to find my way back into ISTE. I’m currently the Chief Innovation Officer and responsible for everything that we are trying to figure out. Everything that’s new, everything that’s forward-facing falls into my area.
Kecia: Which is a ton. And I can’t tell you how thankful I am for you being in that role at ISTE, because the things that have come out of your work, the credentialing, the courses have all been for the good of the whole education ecosystem and really making sure that we’re doing the right thing with technology in a world where we can. . . very easily be accused of doing the wrong thing with technology in schools.
And many schools and districts and ministries are struggling with that. So it’s a never-ending challenge. And thank you for making those resources available and at the ready for us.
Joseph: One thing I want to make sure I’m clear about is that I am the Chief Innovation Officer of both ISTE and ASCD.
The two organizations merged, and while our conversation today is going to be a lot about technology, I want to make sure everyone understands that those are now one organization. The reason we merged is because ASCD is the premier association around pedagogy, and ISTE is the premier association around EdTech.
We felt like these two conversations should be one conversation. And so we very intentionally brought them together.
Kecia: And we can’t have one without the other. Can we?
Joseph: Exactly.
Kecia: That’s been the rub for all of these years. And that’s what’s bringing us to the conversation today around AI.
I don’t believe there’s going to be a day . . . like Pandora’s box has opened and AI has come out of it. And I don’t think there’s going to ever be a day where we’re going to put that back in the box. So it’s the same thing, pedagogy and technology coming together. AI and the world. They’re always going to be together.
It’s always going to be a thing. And I have a presentation that I do that talks about how long AI has been around and it’s not new. We’re talking about it like it just showed up, but it’s been around a really long time. It’s just iterated over time and we’re seeing it in a more transformational state than what it has previously been in.
I know you’re engaged in a couple of initiatives globally and nationally that I’d love for you to share. You’re on an AI task force, which seems like a pretty overwhelming task. You’ve traveled to other countries to see what they’re doing and how they’re grappling with AI.
So can you just spend some time sharing what you’re doing, what your role is or what that task force role is and what are your learnings and then tell us a little bit about your travels and what you’ve learned from other countries.
Joseph: Sure. What you’re referring to is what’s called Teach AI.
We are really focused on how we can convene a conversation around AI and teaching and learning. We’re focused on policy guidance for districts and states. And then we’re thinking a lot about what new literacies need to be brought to schools because of AI. And then last, we’re talking about how education needs to fundamentally change with AI becoming part of the fabric of our daily lives. So it’s a very ambitious organization. We are partners with code.org, ETS—the testing service—and Khan Academy.
And then we also work with the World Economic Forum. As a result, we get industry perspective, the perspective of these giant assessment organizations, and the international perspective from the World Economic Forum.
Kecia: So incredible. A lot has come out from TeachAI.org, which
is a wonderful resource that I reference often to help districts navigate this topic, especially as they’re thinking about policies and procedures and what they need to do to cover themselves around implementing AI in their districts, or even purchasing software that may have AI incorporated in it.
But what other things might Teach AI do that we might not be aware of?
Joseph: Yeah. So just to emphasize what you already said, because I think these are some of the most useful resources out there is we have a policy toolkit. And not only does it give you the areas you need to think about in creating AI policy, but it gives you sample language and examples of schools and districts that have done a good job. It’s a fantastic starting point for you to dive into what your AI policy needs to be at your school. What is coming that people don’t know as much about is we’re now turning our attention to AI literacies, not just in the U.S., but on a global scale. We’re starting to work with international organizations.
I can’t remember if it’s been announced yet, so I won’t name them. But some significant international organizations on thinking about what the global literacies should be for learners in an AI world. This work will have implications on new curriculum in every part of the globe.
It’s a weighty responsibility, and one that I spent a lot of time thinking about. Because it’s easy to make the mistake of focusing on the tools or becoming adept in using an AI tool when really we need to be focused on the principles of using AI tools that will be stable as the tools themselves change over time.
Kecia: Just like with any other technology, right?
Joseph: Exactly.
Kecia: We talk like AI is. . . like we’ve never seen anything like it before. But when I’m talking on the subject, I’ll often reference that this sounds like when the internet came in or when we had email. We were all stunned. What? We’re going to send messages just by typing in this white box and then it’s gonna go and somebody’s gonna send back and where is it gonna go? Who’s gonna see it? What’s happened? How many people are gonna be able to access it along the way? Those were real critical challenges that people have with just implementing email.
Those of you that aren’t as old as me, you don’t remember this. Okay, you don’t need to be as old as me.
Joseph: Sometimes the way I frame it is, if you think of AI as entirely new, then you’re missing the boat. And you think of it as entirely like everything you’ve already seen, you’re also missing the boat.
So it’s not entirely new. All the policies and responsibilities and ways of thinking about technology apply to AI. We still need to be good digital citizens. We still need to exercise discipline. We still need to figure out the right ways that it fits. And we have great guidance already around privacy and appropriate use that can be leveraged for AI.
But AI also has some really powerful new capabilities. And so if you think that the assessments you’ve been using forever are going to be just fine, they’re not. It is going to change some things fundamentally.
Kecia: One of the things that I think we’ve noticed—maybe not everybody, but people in the research landscape are beginning to notice—is the bias.
We need to recognize that and not forget the human element. And when we’re using software that is powered by AI, that there’s a human intelligence that comes by a highly qualified teacher that’s paired with that.
That’s where the power is: a highly qualified teacher partnered with or paired with a very high qualified or high quality AI tool. Those two together can be mind blowing.
Joseph: Absolutely. And I read somewhere where somebody said that you should think of AI as your drunk intern.
And I’m like maybe that was v1 of AI. But it’s still your intern, right? And so if you think about it. . .
Kecia: Your grad student.
Joseph: Like it’s super smart, super eager, knows lots of things, is pretty much up to date, but it has no idea what you are trying to accomplish.
It can just give you input and perspective, and ideas and support. And it’s so good at that. One of the things that we might want to think about is if you approach AI and think, “How is AI going to replace X?” Whether you’re talking about a teacher or just saying, how’s it gonna replace my lesson plan?
That’s probably not as productive a use of AI in its current state than to say, “What parts of what I’m trying to accomplish, can it help me do better or faster or in a more transformative way?
Kecia: I use it a lot in writing and I love it for writing. I think it’s a phenomenal tool to accelerate my writing ability and to improve my writing ability.
It helps me get my thoughts out there faster and more efficiently and organized. After I’ve written, it helps make the corrections or recommendations for corrections that are needed. And I think it’s much improved with an AI tool.
Joseph: I just wrote a blog post today and when I was ” done with it”, I put it into AI and said, “How could this be better?”
And it had great suggestions. It said more specific examples, better transitions. This point you made here, you don’t follow up on, right? Fantastic feedback.
Kecia: Yeah.
Joseph: Yeah.
Kecia: Like a good editor. Yeah. It’s great.
So, what might be the landmines, so to speak, that could be out there with AI? Obviously landmines are hidden and we might not know what is out there. But from your work with AI and what you’ve seen as you’ve traveled the world, what do you think we should be looking out for?
Joseph: I think the biggest hidden landmine is the potential for us to focus exclusively on the efficiency and not actually use it to make our work better. It’s so easy to fall into faster and not better uses of AI. It will absolutely accelerate your work. But it doesn’t, in and of itself, make your work better.
And so the biggest landmine is that we’ll accelerate mediocre practices and be very proud and feel like we’re using AI extensively. We’ll show it off to everyone. We’ll talk about how we’re doing our job in 50 percent faster time. And we’re using it to create worksheets really fast or drill and practice really fast, or one-size-fits-all really fast.
And we’re not actually using it to its full potential to personalize and differentiate and improve what we’re doing.
Kecia: That’s a good one. Man, if only the world will wake up and use it in this way. Is there anything that you can think of?
Joseph: So, one, we are trying to put our . . . sort of money where our mouth is. We’ve developed our own LLM called Stretch AI. If you think about an LLM, it sucks in the entire swamp of the internet, and then you have to clean that out for the stuff that you want it for, right?
No offense to LLMs, right? They’re still amazing, but that’s why you get the bias in there, right? ‘Cause they’re just sucking in all of human conversation. So when we created Stretch AI, we only put in research-based, evidence-based, practitioner-tested resources. So, when you go to Stretch AI and ask it a question, it’s not trying to filter out all the stuff because there’s only good stuff in it.
And so you can ask it, “How can I increase the engagement of fourth graders in my math lessons?” And it’s going to give you research-based pedagogical suggestions for that. One of the things that we haven’t done a lot of is used AI to level us up, but that requires us creating tools that have been thoughtfully designed to help us do that.
And so they’re still a little bit scarce. But I think that’s one major one, I think another one that’s coming is the AI agent. And this is the sort of thing that you think, Oh, I’m not going to do this. This isn’t going to be me. I guarantee you, this is going to be you. So, an AI agent is an AI bot that you train on you. So, it learns your writing style by reading all your email. It learns where you want to travel and how you want to travel.
Whether you’re trying to get points on a particular hotel brand or whether you like to sleep on the top floor or the bottom floor. It will learn all your shopping preferences. All these things. And then rather than you going out on the internet to do things, you’ll send your bot out there to do things.
And we’re like, I’m never going to let AI read my email. And I’m like, if you’ve got a Google account, you’re already letting AI read your email. It’s already being read. Sorry guys. And actually with your permission. That’s both terrifying and exciting because if you really did have an assistant that really was a reflection of you, you really could do things a lot more efficiently.
Kecia: That would be efficient. It’s like Rosie the Robotin The Jetsons. I would like that.
Joseph: Exactly. And it’s also just gonna make it more useful. If when it spits out a sample email to a parent and it sounds just like me, then I have to edit it less.
Kecia: Yeah. Or it can put things in a shopping list and I just edit the list. I would love that. I have to spend a lot of time redoing my Kroger shopping list. COVID changed the way I do shopping anyway. I feel like I saved money by shopping online and having it delivered. So spoiled, spoiled Kecia there.
Joseph: I really hope that we will make intelligent data dashboards for student learning.
Kecia: Yes.
Joseph: Where a parent can always know not only how their student’s doing, but imagine if the AI gave them suggestions for things they could do at home, to reinforce whatever their students struggled with that day, right?
Think how powerful that would be, just little activities that you could do at home to reinforce what they learned or to help them overcome whatever they’re struggling with. That’s not that far away. But again, it takes a really thoughtful implementation.
Kecia: We worked with New York City Department of Ed during COVID.
And one of the things that we did was these parent cards that, for every single unit of every single grade, every single subject, we wrote what a parent could do to help their kid with that unit. We had questions you can ask. Extensions you can do.
These are activities that you can do. This is how it can apply to daily life. Like, I’m measuring how many sheets of toilet paper makes a foot? And you don’t necessarily have to have a ruler to do it. We were trying to make it really easy for families.
If AI could do that work in a just-in-time way, the amount of learning that could be taking place and the amount of growth that kids could be experiencing could be exponential. It would be a game changer for many kids.
Joseph: And I think, you’ve done all that work and it’s probably not being utilized as much as it was during the pandemic, right? If you can train AI on it, it could bring it right back.
I had the privilege of visiting Tokyo, Japan, and Seoul, Korea, where I met with AI innovators. I also met with the Ministry of Education. I really saw some interesting things that I would love to talk about.
I noticed that the educators there were using AI in support of all ages and subject matter areas. This isn’t just math and literacy, but also music, art, even maker spaces. This idea that AI may have less application in the arts was really blown apart by my experience there.
And, to give you a really provocative example, there was an art teacher for elementary kids who uploaded the students’ art into AI for feedback. The students said the AI gave them better feedback than their own parents, because the parents would say things like, “Oh, that doesn’t look like a tree” or “You left the fourth window out of our house when you drew it.” The AI would say, “It looks like you’re expressing how much you love the park”, which is exactly what the child was trying to do. And they found that with AI encouragement, young children were more likely to create more art than with parental feedback.
So, that is awesome. A little weird, right? You don’t think of AI as nurturing creativity, but it’s actually better at it than their own parents.
Crazy. Another example that really struck me is the makerspace person, right? I’m like, how do you use AI in a makerspace, right? That’s the most hands-on place you can imagine. He said that students really struggled with brainstorming. They could come up with two or three ideas when he wanted to come up with 30 ideas.
And so he deployed AI as a companion. A brainstorming companion to them. AI would come up with ideas in the brainstorming sessions, and then the students would build off those ideas. In the end, the final idea was clearly the student’s idea, but it had so much more depth. They’d explored so many more possibilities because AI had prompted them to think in new ways.
And once again, the students reported that they felt more creative with AI present than without AI present, which again, undermines this idea that AI is going to ruin our creativity.
So, I think the last thing that I’ll mention is on top of all of that, they were getting efficiencies of 30 to 300 percent on tasks that they applied AI to.
I think what makes AI super compelling is it not only can be transformative, but it also increases efficiency. How many technologies have you used, especially in a school, that was both way faster and better? They’re either faster and fine, or they’re better and take forever to engage.
And this is doing both. That, I think, really pointed to a future where AI is going to be fully integrated in teaching and learning.
Kecia: It seems like the countries you visited were . . . although many times when we think about Japan and Korea, we think of them as not being as progressive. And especially when it comes to technology, the perception is they tend to hold back a little bit, but it feels like with AI, they’ve really accelerated and said all in, or at least we’re going to try it and we’re not going to create the barriers that we might’ve created in the past with technologies.
Does that feel like what you saw?
Joseph: Oh, sure. And Korea is one, is all in. In my meeting with the Ministry of Education, they shared they are creating what they call AI digital textbooks for every subject matter area in every grade that they will deploy over the next three years. There’s no more all in than that.
Kecia: No.
Joseph: Led at the Ministry level.
Kecia: Wow.
Joseph: And the first ones are going to roll out in March of next year. It’s rule-based AI. It’s not generative AI based. But it’s adaptive AI. Their motto is, 5 million custom textbooks for 5 million students.
So, they’re really trying to say, can we personalize for every single student?
I think another thing that I saw over there, and this really struck me that, to date, if you wanted highly personalized EdTech instruction, you had to go and buy a system that was super sophisticated. And then you as a teacher wrapped yourself around that personalization system. But I saw teachers using AI tools where they can set up the content that they wanted to be personalized and then the system would personalize it.
Kecia: Oh, wow.
Joseph: And so because they didn’t have to go out to a third party, they could wrap it around their teaching rather than the opposite.
And I should say that google.org, the Google Foundation was kind enough to host me on this trip. I was watching teachers who were using Google tools, Gemini and another one called Problem Sets that allow them to do that. They said because they could control it and because students did it ahead of class and because the AI adapted until the students understood it, when they got to class, they had much deeper conversations. And I’m like, this is what we’ve wanted all along, right?
Just like you said, this is what we’ve all been shooting for, but now the teachers don’t have to buy a giant system to do this.
Kecia: And if they wanted a system, they could have it, but that could be their starting place, which is still great. Giving the teacher the opportunity to have the choices
is where the power is. It’s like a physician that has all the different types of medications that you can prescribe or all the different ways that you can attend to a patient and you can pluck them out and say, these are the things that we need to do.
We haven’t had a lot of variety in how we can prescribe learning for students and we’re beginning to get that. And AI is a big part of that.
Joseph: Yeah, I think about my own son who is extremely passionate about wolf conservation, and there was a time when he tried to make every middle school assignment about wolves, no matter what the assignment was actually about.
And it was a little challenging for his teachers. But there’s no reason with AI that he couldn’t be doing wolf conservation math and wolf conservation language arts, and somebody else couldn’t be doing ballet math and ballet language arts, right? So, there’s a cool future where we can hyper personalize to students’ interests that I think is really exciting.
Kecia: Yeah. And it’s really exciting for those kids too, because they may learn concepts that they never would have learned otherwise. I give a story often where when I was learning math, I had a math teacher and he taught with sports analogies all the time. And I wasn’t that into sports. I had no idea what he was talking about, probably 80 percent of the time.
When I went to college, who did I see in my freshman college class but that dude that was my senior math teacher. He showed up at the college. I was like, are you just on the planet to torture me? Is that your purpose in life?
I dropped his class immediately, transferred colleges because I never wanted to have him again, and it created such a math phobia that I ended up getting a minor in physics to take care of all my math credits.
Joseph: This is such a great example. And I don’t know what you were into at the time, but had somebody been able to relate math to that. . .
Kecia: Gardening.
Joseph: . . . almost every concept could have been related to gardening and it would’ve changed everything for you. Richard talks about how he couldn’t learn any math until he wanted to become a pilot, and all of a sudden he was doing all the math that he hated because he wanted to fly.
And it was relevant and interesting. And I think the potential for AI to bring relevance is really unlimited.
Kecia: I’m really grateful that you’ve taken the time and shared all this wonderful information with our audience.
I think you’re a treasure in our industry and a gift to educators everywhere. They have no idea of the work that you do for them. And I just want to say thank you on their behalf for everything that you do to make their life easier every day.
Joseph: This has been a joy.
Kecia: I appreciate the work that you’re doing now, investigating how AI can be so helpful. You’ve offered great suggestions and it’s going to be fun to come back in two or three years, and it’s going to all be happening.
Joseph: Yeah, that’s my big caveat all the time is this technology is moving so fast.
Kecia: Yeah. It won’t take two or three years, will it?
Joseph: None of us are going to be able to know exactly what’s going to happen. We’ll talk in six months.
Kecia: Six months. Talk to you later.
***
Prepare students for lifelong literacy with Writable, a program designed to help students in Grades 3–12 become proficient writers through AI-powered writing feedback and daily practice.
