AI Ends Knowledge Scarcity — Now the Classroom's Greatest Asset Is Human Interaction: AIxHigher Ed March 30
AI is changing the student learning experience by collapsing knowledge scarcity — students no longer depend on a single teacher, textbook, or institution to learn, since AI tutors can teach virtually anything, even the world’s best content – anytime, with personalized, one-on-one instruction at scale.
This shift, building on earlier transitions from oral to written to internet-based learning, means the classroom is no longer the primary place students acquire, or need to acquire, content. As a result, the role of the teacher must evolve from “sage on the stage” delivering knowledge from one head to the next to something more like a coach or learning architect — someone whose focus is on building relationships, modeling critical thinking, facilitating collaboration and debate, and helping students develop the human skills (motivation, teamwork, conflict resolution) that AI cannot replace and that students will need more than ever in an AI-powered world.
Chapter 1: Why This Conversation Matters
Anand Rao: In one of the previous episodes, Stefan and I were talking a bit about how AI is not only impacting what’s happening in the classroom, but because of students’ interaction with AI and different technologies outside of the classroom, many of us in the classroom aren’t really fully aware of how it’s impacting our interactions with students.
So, we wanted to develop this as a separate discussion, as a separate episode to really look at the student learning experience.
The way that AI is transforming so much of the world. We need to be aware of how it’s impacting their access to information, the way they learn information, the way they have different interactions with other forms of data that it will impact their expectations in the classroom.
So when we start going from an oral tradition to more AI-powered tutors, we need to be aware of how that changes the cognitive process for our students, changes their expectations as well as our own expectations for what they could be doing.
Stefan: If you look briefly in history—and you know there may be more of these—but, well, how did students learn things? If you went back originally, whether it was a tutor or a lecture, students learned of orally directly from one person to another. You would teach me stuff and that was basically the only way I could learn.
Of course, then we had the written era, the invention of the printing press. People could learn stuff from books actually because you know, maybe wars, right? A lot of wars were fought throughout the world because of the way kind of knowledge shifted.
Now, of course, even in the book era, like you had to go to the library to get the books, right? And we both kind of transition through that. Like when I was in high school as a debater, I had to always go to the library to do research.
But then we were in college, you had the internet era, which opened everything, right? Because we both grew up in Erie, PA, and the biggest libraries were kind of small. I remember going to the Michigan debate camp and I was like, “Oh, we’ll use our library. It’s huge, right? It has all these resources.”
But the you could get any information you wanted online and people built what were called MOOCs, which were, you know, kind of on—I forget what the acronym stands for, but open online communities.
Anand Rao: Yeah.
Stefan: Where you could learn like all these things, but you had to go actually—you kind of had to learn it yourself, right? Nobody knew—the MOOC couldn’t teach it to you. You could listen to a lecture. A lot of people recorded lectures and put them up and it was a way of learning. It didn’t work out to be the same way that everybody thought it would work. It didn’t, you know, replace all the teachers.
Chapter 2: The Evolution of Learning: Oral → Print → Internet → AI
Stefan: But we have the AI era. In the AI era—and we’ll talk about this a little more—you know, it’s different in some respects. The AI is interactive. It can teach you things. It can teach just about anything. It’s all in one place. Like even with the MOOC, I’d have to go find my MOOC on X, Y, or Z and then go find another MOOC or find something. Now I can just ask the AI almost anything at this point. It’s going to help me out. It’s going to change that.
And I’ll say the one general theme—and this is what I think prompted us to have this podcast—is that when students approach the classroom, there’s so many different ways they can learn. It’s not just that they can get the knowledge from the teacher; like they could also learn about it in their textbook. Teachers are primarily teaching out of the textbook. They can go look on the internet the old-fashioned way. They can now just on their phone ask the AI and more reliably get answers if it’s a concrete thing or get them to help think through something. So, I think the one thing, you know, at least I want to talk about—not to talk about, but to point out, to highlight—is that students when they’re in school, the classroom is one way they can learn content. And even on our last podcast we talked about how you can learn by texting; you can get a text course about an intro to AI. So the teacher is only one route of information delivery.
Chapter 3: Supply vs. Demand Learning
Anand Rao: Another aspect of this that’s really important that we want to tease out is the way that information or knowledge is supply-driven or demand-driven. If you went from the oral era to the written era and you were thinking about if you heard an idea, you had to memorize it. You had to think about it. When you get to the written era, you’re like, “Well, I know that idea is out there. I know it’s in that book. I can go and find it when I need it.“ And so I can access it. I don’t have to memorize it. And that was one of the worries originally of moving out of the oral tradition is that people would lose their memory. They wouldn’t have the strong memory they had before because they knew they could recall that information by going and looking it up and finding it somewhere.
When you get to the internet era, it’s about accessing still and you go and seek it out, but you have to be active in seeking it out.
An aspect of this when we get to the AI era is that similar to maybe in the classroom, it’s supply-driven. It’s the instructor trying to supply the information to the student. In the AI era, we’re going to work into a model where the AI is able to not only offer it when asked, but the AI might also prompt somebody to say, “Hey, I see you’re doing this. Why don’t you try this?” So, it’s kind of like the advanced Clippy. “Hey, I see you’re trying to write an email. Would you like some assistance with this?”
We’re now seeing something that’s far more capable, that could be more subtle and certainly could be more helpful as it sees somebody struggling with something or sees there’s a problem, or notices that a student is spending more time on a problem than maybe they did before to be able to jump in and provide that knowledge. This is really going to change then; if students aren’t thinking about it as being demand-driven—which is really where we start thinking about intrinsic motivation, they’re looking to be able to better themselves—if they start to rely on something that’s supply-driven and it meets all of their needs, how does that affect their motivation? What would that do to intrinsic motivation? And that’s something that we just don’t know yet. We’re going to have to play that out.
Chapter 4: The End of Knowledge Scarcity
Stefan: It’s just a fundamentally different world.
We have a world where knowledge is abundant, which is another thing we could talk about. You could do a whole podcast on that.
Originally it was only in a limited number of people’s brains and then it had to be shared with the other person.
And then it became in books, so it kind of declined in scarcity that way. But then universities in some way really kind of protected the scarcity. They kind of said, “Look, if you want to get...”
You know, every university claimed they were the best, but a lot of people think Harvard is the number one, so let’s just use that example.
And you say, “Okay, well, in order to get—Harvard has this special knowledge and only so many people can get it, right?“ You gotta really kind of kill yourself in high school. You gotta jump through all these hoops. You gotta—and this really isn’t true, but theoretically—it’s like, well, you got to get 1600 on the SAT. You got to get all fives on your APs. You got to play sports. You got to cure cancer. You got to start a nonprofit. I mean, I’m being hyperbolic, but you know, you had to do like all these great things to get into this place that had like this kind of special knowledge that was kind of limited in supply, and now everybody can get it.
It can be personal, can be adaptive. You were talking about that a little bit indirectly like just a minute ago, right? Like it sees where you’re struggling or what you’re doing or where you’re spending more time on. Though it becomes personal, then it can move and adapt that instruction.
That’s got an adaptive learning system—which are, you know, they’re still kind of in the early stages, we may have some instances where they’re working pretty well—and then always available, right? Like regardless of whether you want to go to Harvard or somewhere, the professors aren’t always available. But the AI is always available. If you wake up at 3 a.m. and you can’t sleep and you decide that’s a good time for you to learn, you can learn. So, these are huge changes. And you know, but maybe it’s 2 p.m. and you’re in class and you’re not feeling like paying attention. Well, that seems bad. The teacher gets frustrated. The teacher’s thinking, “Well, if you don’t pay attention, you’re not going to learn this.” And you’re like, “Well, no, I’ll go learn it at 3:00 a.m. when I wake up. I’ll still learn it.”
But you know, there’s a—maybe I will or I won’t, right? I mean, that obviously happens, right? And that kind of goes back to the motivation, but it’s not like the only place I can learn this is in this classroom.
Chapter 5: The Collapse of Credentialing?
Anand Rao: And even in that regard, it’s something that is a profound difference. But if you look at some of the most elite universities, that has already shifted over the last generation or so. It’s not like when we have things online, the only place you could get it was going to Harvard or Oxford. That kind of shifted those roles; then they were more about credentialing and being able to say, “Well, we’ve identified that these are superior students.” And so even if they never set foot in the library, they’ve been identified and they get that credential.
Now we’re moving on to something that is compounding not only the loss of the scarcity of knowledge because it’s now abundant—it’s generally available to everybody and it’ll be even more available and it’s always on—but we’re also moving away from a credentialing model where people can just act on their own and it’s really more about being able to develop their own skills. That changes the motivation piece as well and that factors into what happens in the classroom.
Stefan: Yeah. And you know to kind of the point there —when these MOOCs and stuff came and then people just started putting courses online—I think MIT put like all or nearly all their courses online. People said, “Oh my gosh, no one will want to go to MIT anymore.” But you couldn’t get their credential that way and everybody still wanted the credential. But as you’re talking about, the credentialing is starting to collapse. As long as you can do it, a company will hire you and you can demonstrate that you can do it. And if you fake that you can do it, you’re not going to last anyhow. So that, you know, it’s a little short-term harm to the company. But it doesn’t matter; the value of the credential is declining.
Another big thing is one-on-one teaching at scale, right? There’s some research on this, but I was telling somebody today, “You can go back and forth about the research, but it’s just kind of intuitive.”
Chapter 6: One-on-One Tutoring at Scale
Stefan: Like, if I sit down with you—and math’s kind of the easy example because you learn it linearly, things build on one another. And it’s objective and you can assess it. If I sit down with you and I’m just like, “Okay, well, what do you know about this concept?” And we talk about it and I can use examples that are relevant to you and I see where you start to fade and where you don’t understand, and we don’t move on to the next thing until you understand it—which does not happen in the classroom. You could have five students fail that whole thing, math, whatever the teacher is trying to assess on that unit or chapter test, so to speak. The class moves on, you don’t understand it.
So obviously when you’re teaching one-on-one, you should be able to move faster and the students should be able to learn more. It’s just obvious and there was no way to scale that until we had AI, at least in a cost-effective way.
Anand Rao: The initial attempt to do that was the flipping of the classroom and it was the move with Khan Academy to be able to say, “Watch the lecture at home and then when you come in you can get individualized attention,” and the model that we should make sure that people are mastering content before they move on to the new content. But it ultimately wasn’t practical because we didn’t have the staffing. We didn’t have the infrastructure for that to be available. We didn’t have people accessing it on their own sufficiently. And so it didn’t really meet—it didn’t come to fruition. It didn’t work out the way really we would have liked it. So we’ve identified this as a problem for a long time and we just haven’t had the means to be able to really address it.
Now think about a couple of implications to this. The first is not only is it a personalized tutor, but as we’ve talked about in previous episodes, AI is going to be capable—and already is capable—of being hyper-specific and persuasive. So, it’s not just a matter of having a personalized tutor, because a personal tutor could just meet with any individual one after another and teach them through that concept. This is a personal tutor that knows you so well that it can really use the exact examples you would need to use and teach it in the way through the modalities and through the practices that would be most effective for you.
Chapter 7: The Rise of Hyper-Personalized AI Tutors
Anand Rao: And of course that’s also part of the concern we have about hyper-persuasion because it’s able to use those techniques so well. But that’s an aspect of this—that it’s what happens when the students get so used to a personal AI tutor that knows them that well, that is so specific and that supportive and that effective. How many of them are going to say, “I’d rather work with the tutor that’s AI than a teacher“? And that might be a concern; that might be something we start to think about.
The other aspect of this is if we have access to that, and if students have access to that at home and they come to the classroom, if we don’t change what’s happening in the classroom, they’re going to be disengaged. Why would they want to just listen to the old “sage on the stage” giving a lecture on something when they know they could get that information from another source—their personalized AI tutor or even just from a video that they’re able to watch at whatever speed that they wish and maybe one that’s already been personalized for them?
It’s not just a matter of this happening elsewhere; we have to think about how that seeps in to the classroom. What does that mean for the way we need to change the way we teach so that even if we’re not employing AI personal tutors in the classroom, we better reform what’s happening in the classroom? Because the reality is students will expect that outside. And if we don’t meet that need, if we don’t meet that expectation, then we’re really going to be failing them and they’re certainly going to be disengaged.
Chapter 8: AI as Curriculum Architect
Anand Rao: And here we get into this discussion is AI as a tutor, a study buddy, but also this idea of a curriculum architect. And I really like the way this gets teased out and the way Stefan put this in the slide deck and leading to what it means for a teacher as well. Think about whoever would direct what happens and the person that builds and designs what that curriculum would look like.
Stefan: It’s just like kind of getting built out and adapted to a small set of students or an individual set of students. It can be the builder and in many cases, right, you have teachers who are not curriculum designers. And over the last decade, we really kind of seen an explosion in curriculum design. And I noticed it when I’d taken a class that was designed by a designer. I was like, “Wow, this is an amazing class, right?” And it kind of hit me. It did help me learn. But the AI can be trained on curriculum design. And it’s getting better and better. And you see even curriculum designers saying, “Wow, this is like getting really good.”
So, it maximizes that as well. And this just kind of reinforces kind of the point like you know I was saying earlier that learning is no longer confined to the classroom. And I don’t know how to emphasize that enough. Right? Whatever it is you need to learn to kind of get this grade in school, the curriculum slash the school slash the teacher—they kind of need to set that for you to earn that “A.” They don’t necessarily have to, but if you want an “A” in that class in that school, then you’ve got to kind of meet the curriculum requirements that have been established.
Chapter 9: Learning Beyond the Classroom
Stefan: Unless there’s a lot of interactive element in the classroom that that’s also being graded and assessed—and we’ll talk about that later. If it’s content, then you know you can learn that in so many places.
Anand Rao: And the other aspect of this that’s really important to note is look at some of the subcategories here. These aren’t just about personalized AI tutor bots. These are developments that have been occurring over the last couple of years and most of our students are already accustomed to “EduTok” as a phenomenon. People are using TikTok and using that platform to be able to share really short science explanations or discussions about history or about government and civil society. There’s a lot that’s being imparted to students that way really effectively. Same thing with shorts on YouTube or Instagram Reels. It’s not just about some influencer or some entertainment, but there’s a lot of good quality content that’s there.
And at the very least, we need to recognize that that’s been building for a while. Our students are accustomed to it and, you know, there are concerns about what that might do to attention spans. I think for all of us that’s certainly impacting it, but we also need to recognize that they have access to information other places, which means there’s opportunity for us to be able to share information in other places and to integrate that into what goes on into the classroom. We have to break down the idea that only what happens in the classroom is what’s important. We need to recognize it’s happening outside and we need to find a way to bring that back into what’s happening in our classroom with our students.
Stefan: And just kind of two comments there. I think it was a video I was watching and it was definitely CMU, Carnegie Mellon—you know, one of the best universities especially for science in the world. The professors—I can’t remember if it was math or coding—but they were teaching math students through Twitch. A short kind of video segment platform. And they’re using that to teach these really advanced kind of undergraduate and graduate students math.
I think the second thing is, you know, I think about—I’ve been thinking a lot about debate because I think it’s always—I thought it was important, you know, spent my whole life in it and I think it’s kind of really important in this AI era. And we got to think about ways we can scale instruction, right? Like we talked about in the podcast we did on the update before, about how the government is scaling AI literacy through text that everybody engages in. The people who want to get the word out, so to speak, who want people doing stuff that they think is important, they need to start using these learning methods and not just relying on human teachers delivering the knowledge in the classroom. Even if you, for some reason, think that’s better, it’s not scalable.
Anand Rao: The other part of this is really getting to that third column. Content isn’t everything. And I think this is really the conclusion we need to start to draw is that, you know, the teacher is not the only option. Students have genuine choice and we should recognize that we have choice in how we’re designing this. If we recognize the opportunity for students to be able to learn outside of the classroom with personalized tutor bots, with EduTok, with content that’s online, and more and more with devices that are incorporating personalized agents that are going to give them access to what they want to know, what they need to know, and be able to do in a personalized way..
Chapter 10: The New Purpose of the Classroom
Anand Rao: ...then we can get to the stage where we can make the most use out of what’s going on in the classroom, and that is about human interaction. Every time I hear “human in the loop,” I’m thinking what that really means is we want to make sure that humans are still putting themselves center at what is important for humanity, what’s important for society. And how do we do that? If we’re all only working with AI tutor bots, we’re only working with agents, then we’ve kind of lost sight of that.
The classroom should be the place for relationships, for teamwork, collaboration, projects, project-based learning, debate, discussion, and we need to model this and we need to be able to provide this space for students to be able to do it. And if we don’t teach them how to do it, then they’re going to be really ill-equipped to be able to take that on beyond their education. And this is a really important—I think this is an imperative. We need to recognize the role that AI can play in helping with their education, but we need to do more than we’ve ever done in previous generations to teach social skills, to model social interaction and collaboration.
And we need to be able to think about really the tough cases where they have friction, where they have disagreement, where they have conflict, and allow for that conflict to occur so they figure out how to manage it. Because if they don’t, and they run away from it, then they’re going to lose that human connection later on. And I think that’s more of what this means for the classroom than anything else. I certainly see all of this happening with artificial intelligence and the way it’s going to be able to be a real benefit for education as well as some concerns. But I think that’s going to play out. But what role we have is a determination of what we do with our human interaction with our students and for our students to interact with one another—that we can be thoughtful about and we can be very intentional about, and that’s the human connection.
Stefan: And you know, that’s really why it’s empowering to educators. A lot of educators are looking at AI and saying, “Oh, well...” They kind of have two concerns. They’re kind of in tension with one another. People can hold two ideas in tension with one another. That’s fine. But the two ideas that are in tension are: one, “I’m going to be replaced by an AI,” and two, “AI just can’t do what I do.” You can’t really believe both of those at the same time, right?
But you got to think then about what it is that you can’t do. It’s unlikely that you know more content than the AI, right? Unless you’re like at the very top of your field—but even the top mathematicians are saying it can do some math better than I can. So it’s unlikely the content. But then there’s the question of, “Okay, well there’s the teaching of the content, right? In kind of getting the content from your head into my head.” And some people think, “Well, I just must be better than that. The kids want this human interaction, right? Like just like this lovely human interaction.” But you know, that obviously is a little bit debatable because I guess we have to take away their phones in order to get them to interact with us. So maybe they do or they don’t want the human interaction.
Chapter 11: Teacher as Coach, Not Lecturer
Stefan: But the point—like what you were just saying—is: well, do they want the human interaction? It’s not just about like delivering the content. This is not the only place that can occur. It’s like if you’re on a sports team, right? It’s not just the coach telling you something. It’s the human interaction. It’s kind of the working, the teamwork, right? All those types of things. Empowering you, encouraging you, doing something as a team that you couldn’t do as an individual.
And we talked about this from the very beginning. And your son talked about this from the very beginning, right? “Teacher is coach.” You can go find the YouTube thing from the conference—I think from 2023 or 2024, the latest it was 23. Yeah, I think it was August ‘23, right? At that online conference, talking about his experience in debate and kind of the teacher as the coach and now he thinks his future of education.
It’s hard for me—I try to be as open-minded as I can—and it’s hard for me to wrap my head around any other role of the teacher other than kind of the coach, which I think is a great role. I’ve been a debate coach as you know my whole life. And I think it’s a great way for teaching and learning to occur.
Anand Rao: And I also recognize, though, that that doesn’t come naturally to everyone or that we’ve necessarily been trained or coached ourselves to be in that coaching role. It can be a little jarring, but what I love about the message on this slide is that it can really redirect you back toward what drew you into teaching in the first place. And I think for virtually all of us, that’s the interaction with the students. If you didn’t want to interact with students, you wouldn’t be a teacher.
And so think about it in terms of: “What can I do to support them, to guide them?“ Instead of having to just impart knowledge that can be done with other tools now and can take away a lot of the drudgery that’s there. And we want to be able to instill in them intrinsic motivation, critical thinking, collaboration, and thinking about what it means to be a human in a world that’s powered by AI. That’s a lot of discussion and that’s a lot of interaction that we can have with our students and that we can encourage them to have with one another. I think that’s really empowering.
And so it might be a little bit scary at first, but I think we have to consider that education is going to change and it needs to change. That doesn’t mean it has to change for the worse. It’s going to change for the better if we decide that that’s where we want to take it.
Stefan: And you know, so there are some things, right? It’s like kind of only you—or most likely you—can do. One is kind of to “make” students care. And I don’t know if I love that word “make” in there; you can’t really force anyone to care, right? But you can help them care. You can help them prioritize. You can explain the significance of something. You can get them to where they’re doing an activity where they like it. And you don’t have to expect 100% solvency—every kid’s not going to care. But every kid doesn’t care now. It’s not like every kid can’t wait to get to math class to listen to the teacher write on the board how to do the math problem.
So you’re not going to get 100% solvency in the model, but that’s something you can do. Build real relationships. Like kids who say, “The most important teachers in life” are often also their sports coach or the debate coach. There are other teachers of course that have that impact that don’t have those roles, but that is how you can build real relationships with your students.
Model thinking. Model interactions. To me, that’s what I try to do. You know, I don’t really do that much direct teaching of students unless they’re in “Intro to AI” or “Intro to Debate” or something like that where you do have to kind of build some content knowledge that they don’t have or help them acquire it. But I always try to work with them to improve and have a consultation process. So, you’re kind of modeling that thinking and then: look, recognize when AI misses, okay? And recognize when you miss. And I think, you know, that’s obviously hard. Like, no one wants to recognize when they miss, especially if somebody else tells you that you missed. Nobody enjoys that.
Chapter 12: Teacher as Learning Architect
Stefan: But you know, recognize when everybody and everything misses and kind of improve that situation. And look, the AI is going to be looking in some instances, right? When you have an adapted learning system, they’re going to be looking and seeing where the student misses and where you miss. So but that’s all part of a team. We talk about people and AIs working together as part of a team. That that’s part of the team.
Anand Rao: This gets back to something we were talking about in a previous slide: the idea that AI can be a curriculum architect. But really the teacher—and I think this is a really empowering role and something that I think we can all grow into and build into—is the teacher as the learning architect. We can be the primary source of designing the learning experience. So we don’t have to just be the primary source of content; we can be the primary source of the learning experience. We can design that learning experience.
And think about how much more fun we can have when it’s about those interactions. It’s not just about supplying information, but engaging with those ideas, modeling them for our students, and developing those communication skills, the collaborative skills, the interactive skills that we want them to have. This is higher-order thinking and it’s a higher-order role. It’s not a diminished one; it’s a more important one.
And if we can get rid of a lot of the drudgery and get rid of a lot of the stuff that maybe we don’t enjoy as much—and sometimes it’s fun to give a lecture—but think about turning it into a dialogue and having a discussion instead of just sharing it to them. Having them follow up with questions and having discussions about what you would do with it. As the teacher, you can set the goals. And you should teach students how to set their own goals. We’ve talked about in previous episodes that it’s really important for students to start to develop that understanding of judgment of what their goal really is. If they haven’t determined what that goal is going to be, then it’s really hard to work with Agentic AI. You have to be able to say, “This is where I want to go. This is what I want to produce. This is what I want to work through.”
And if you can set those goals well and you can model it for your students, then they’re going to be able to benefit not only in the classroom, but also in their own learning experience as they go forward. There’s also a matter of choosing the right tool. You’re thinking about the interactions, you’re thinking about how that’s going to be developed. You can do that effectively for students early on and as they get a little bit older, you’re also modeling that so they can start to curate their own ongoing path.
There was a really great article in Forbes recently that talked about how higher education needs to stop thinking about students that come for a degree and then they leave, but having a lifelong interaction and relationship with that student. And that’s really a more healthy interaction. That’s a more healthy relationship that we should have, and that we’re able to continually improve their lives. They can continually give back to other students as alums and as people that have worked through some processes. Part of that is that you want them to be able to be part of that path. You curate it and you teach them how to curate it and be able to incorporate others. And then finally: creating the conditions. It’s about designing these experiences that will help them develop that intrinsic motivation to make sure that learning is going to stick and that they see the purpose—make sure it matters to them.
Chapter 13: Why Education Can’t Stay the Same
Anand Rao: So I think there’s a real empowering role here for instructors, but we have to grapple with the idea that it’s fundamentally changing through our interaction with students, through students’ interaction with technology, and the way society is changing as a result of all of this.
Stefan: Yeah. You know I’ve commented about this before at a broad level. It’s like, I think some people think like we reached the “Francis Fukuyama end of history” book, right? So we had liberal democracy, we’re at the end of history, we’re going to have liberal democracy forever. I feel like some people think we’ve reached the end of education. This convention, this classroom—like even in some optimized form that kind of is less factory model—that’s it, we’re done, right? Regardless of how the world changes, we’re going to have the same model of education. It just doesn’t make any sense to me.
Now, you know, as we discussed on the last podcast, computers are in many ways smarter than we are and are more intellectually capable. So given that education is focused on intellectual development, the idea that what we’re doing doesn’t need to change—at least in some ways that we can figure out in that world—it just kind of boggles my mind. It doesn’t compute.
But you know, here we really kind of have a path towards empowerment: the humans and the AI working together. And look, you can think of it—almost in this instance—you know, someone will yell I’m anthropomorphizing, but just think of the AI as like another smart person that you get to work with, right? And you bring all their strengths into your classroom, and you bring your weaknesses too, just like you every day bring your own strengths and your own weaknesses into your classroom. People say AI’s intelligence is “jagged”—or at least they have said that. Some people are saying it’s really not that jagged anymore, but so is human intelligence. There are many things I wasn’t very good at, or at least I didn’t care to apply myself towards, and I bring that jagged intelligence every time I work with students. Because maybe we get to something—because I didn’t choose to develop my knowledge in a particular area—that I can’t help them when we get to that part of the debate argument or whatever. So everybody brings in their strengths and weaknesses just like AI, and if you have enough entities working together then collectively you do reduce that jaggedness across the group.
Anand Rao: It’s really the appeal to working in groups, to a pluralistic model where you’re able to bring those different perspectives together and you know that the sum of the group is greater than the sum of its parts. It’s about being able to produce something you wouldn’t be able to do on your own.
Chapter 14: What Students Actually Remember
Anand Rao: We can think about doing that with AI as well.
Stefan: And you know this just kind of sums up here, right? That you know students don’t remember the worksheet, right? They remember the teacher who changed how they saw themselves. It’s that inspiration. Don’t get locked up in the content. If you want to have an impact like you say teachers do, then you know, make it that. Some teachers do get excited about teaching their content especially once you get out of the primary and middle school years. People became high school teachers because they really like math or they really like social studies. Obviously professors—”Oh I really like philosophy,” or whatever.
So there is a lot that kind of comes with the content, but I think even when people look back—and I’ve been working with students for a while—it’s really the interactions you remember as a teacher. Not like, “Oh, I really taught that kid how to extend that disadvantage well in the 2NR” in my instance, or “I really showed them how to get that formula.” If you feel so good, it’s because you got them to understand it, it’s because they learned that formula.. And that’s kind of the key role.
Anand Rao: And the last thing I would say about this, as another additional reason why this should be appealing and we should lean into this, is that that human interaction is something our students are going to need more than ever. Because if we are concerned about AI companionship and students over-relying on chatbots or becoming less interactive and not developing social skills..
Chapter 15: The Risk of Losing Human Skills
Anand Rao: ...then it’s imperative that we find the space, that we provide the opportunity and we model for our students how to do that effectively. How to have that human connection and how to engage with one another to appreciate human connection. Because I think for some, we should be concerned about whether or not that’s going to be there if they take it too far, if they don’t know how to handle that.
Remember, these are students that are developing, and developmentally they aren’t going to be ready to make a lot of those decisions on their own. We can’t leave it all to them; we have to be able to guide them. That’s why the educational process is so important, and that’s the area where I don’t think teachers can really be replaced. Teachers as parents and teachers in the classroom—we need to be able to guide them and to be able to curate that experience for them. It’s just going to be a little different experience than what we experienced growing up because it’s a different world.
Stefan: And of course they’re just apprehensive about their futures, right? Whether you think things are good or bad, instability is always worrisome. So you kind of have a pretty significant war going on, which maybe could end or it could get a lot worse. And you have more and more reports of very well-educated people who got jobs in some of the places that are hardest to get jobs in—top level Amazon leadership roles, blah blah—these aren’t just jobs that anyone can have. When they’re struggling to find work, you may be sitting in class and saying, “Well I could learn this and it still won’t help me at all.” And you know, “I took out a bunch of loans, mortgaged my future.” Like the podcast last week: “Students who mortgage their future as they sat in classes and the jobs disappeared.”
Those are challenges and we need to be supportive of our students. But you know, kind of always looking positive forward, right? It’s not a story about replacement, it’s a story about doing different things and it’s a story about kind of leaning into the shift. Faculty do need to lead in the shift if they want to be relevant.
Chapter 16: Final Thoughts: Leaning Into the Shift
Stefan: We can’t ban AGI. We can pretend we’re banning it, we could take away a phone or a device, but really we’re not going to ban AGI. And that’s the world; whether they have their phone in class or their computer in class, they’re going to live in an AGI world. And we can’t ban that. So, we can’t just make the classroom into a fantasy space if we want it to be relevant. So, we need to kind of lean into this shift and kind of give them beyond content. Like I say, it’s humans.
I think everybody’s starting to really think about, “Okay, well, what’s my value as a human?” And it comes from interaction with other humans. I almost was going to say we don’t have any intrinsic value—but people would have said, “What do you mean?” No, I think you have intrinsic value as a person, but I’m saying in terms of what you might bring economically. We need to all make a certain amount of money. What you bring to the table is going to be not just monetarily, but in terms of your own personal development and growth. The value that you have and that you get, and your positive experience in life—they come from interacting with other humans. I don’t think that you get that from an AI, at least most people aren’t.
Anand Rao: Yeah, there’s a lot more here that we’re going to be exploring and there’s certainly a lot more that is going to be developing. We would love to hear what you all think, certainly in the comments, but you can find our emails on the website – AIxHigher Ed – and you’ll be able to share with us your thoughts on where you think the classroom is changing, how can we contribute, and how do we make this transition so that we can better provide the educational experience for our students that they really need.
All right, great talk, and thanks Stefan.