[Music] We're back with season two of the Humans of Learning Sciences. I'm your host, Dr. Monlin Monacoco. I'm excited to announce that we'll be shaking things up a bit this season. The next episode of season two will feature two guests who work in similar spaces, but have unique perspectives on what it means to design for and support learning. Similar to last season, we'll get to hear the nitty-gritty stories about the different journeys that have led them to where they are today, as well as how their thinking has shifted along the way. Here's a quick sneak peek. We talk to folks about research-pacti-sheno partnerships, civics education, collaborative learning, and so much more. I can't wait to hear about what you learned from these episodes. You can follow us on Twitter at humans.lspod or email us at
[email protected]. [Music] If you've been listening to the news over the last few months, you know that there's been an incredible interest, debate, and lots of speculation about the use of artificial intelligence or AI technologies. I've been really interested in the kind of verbs and metaphors that have been used to describe this relationship between humans and AI, as collaborators, as replacements, or as forms of surveillance, for instance. And I've been really curious about the view that learning scientists take on these emerging technologies. So I'm thrilled today to be bringing together two scholars who've been thinking about this, both the opportunities and the challenges that arise, as AI technologies make their way into schools and classrooms. We'll be talking with doctors Jeremy Rochelle and Thomas M. Philip. Jeremy is the executive director of Learning Sciences Research at Digital Promise, and Thomas is the professor and director of teacher education programs at the University of California Berkeley. Welcome to the podcast, Jeremy and Thomas. Monica, I've been waiting for this moment. I'm really looking forward to our conversation today. Good to be here and thank you for having us, Monica. In doing research for this episode, I realized that both of you were grad students at UC Berkeley. Jeremy graduated in 91 in Thomas in 2007, and that you had both the same advisors as well while you were there. Both of you came in to graduate school with engineering backgrounds in computer science for Jeremy, and for Thomas, computer science, and electrical engineering. How did each of you become interested in a degree in the learning sciences? Jeremy, can you start with you? Sure. Happy to talk about that. So before coming to Berkeley, I was studying computer science at MIT, and I was lucky there to be working with people in the logo group who were studying constructivism, and really cared about kids learning. I felt myself as an undergraduate, so I have to make a choice. Am I more interested in computers, or am I more interested in people? And for me, I was always interested in what computers can do with AI, but because it gave insight into a theory of mind, and it helped me understand people. So I just decided that I didn't really want at that point to go get a PhD in AI, but rather I wanted to double down on understanding people and what technology meant to people. And at the time, Berkeley was launching a new program that would really become a learning sciences program, and that seemed like the ideal place to go to that. So that's how it happened for me. What about for you, Thomas? It's funny, I don't feel I realized I was in the learning sciences until well into the program. My undergraduate degree, as you said, was an electrical engineering computer science, and I think there's a push pull factor. I love the design aspects of engineering, and I was also having a lot of questions about what I wanted to do with an engineering degree. So simultaneously I started volunteering with an education program for the unhoused, which really got me thinking about education and about learning, but also some of the structural inequities in society, which encouraged me to finally decide that I was going to become a high school teacher after I finished my degree in engineering. So I went back to LA and I was teaching in South LA for a few years, and then in back to graduate school, I think as an opportunity, they wore into questions around teaching and learning. And it was only then that I realized that it was entering a field of learning sciences, which was still in some ways very, very nascent at that time. I'm curious about sort of your graduate student career, the kinds of things that kind of shaped the things that you're doing now. What interest did you have? What sort of pulled you into the field as you were going through your graduate career through those introductory courses, and then at the end and thinking about what careers you were going to pursue. Any sort of inflection points that you can think back to or formative experiences that really shaped who you were at that time. Like Thomas, I came from engineering, I didn't know what I was getting into. I remember early on, people were using the term protocol and in engineering, it's really clear what a protocol is. It's like how you communicate down a wire, for example. But here people were actually using it to mean capturing a recording or an interview as you listen to students think. So there was just a lot that was new in the first year of what have I gotten myself into. But two really big inflection points for me. One was I was lucky to be at the time when situated learning was being invented. And it was being invented by a team of people that were organized around Xerox Park, around Stanford and Berkeley, and they got together periodically. And so I was on one side doing a dissertation about how students learn physics with technology. But on the other hand, I got super interested in collaborative learning and the role of gesture and the informal ways people used conversation to make sense together and to come to stronger concepts with each other. And you know, funny thing was that I think looking back now, this was a long time ago, the straightforward work that I did has directed by my Berkeley program just on cognition and mental models of learning physics. That's a lot less cited than the collaborative learning work, which was really a turn, you know, just because what was going on at the time. And I'll just say one other really big turning point, you know, is that that moment where Steve Jobs talks to John Scully and he says, you know, if you want to make sugar water for the rest of your life, or you want to build the next great thinking machine. And that Scully goes on to lead Apple. And a moment for me like that was I came really in really interested in how people learn physics. And that's what my dissertation was about. But sometime later Jim Caput, who was a math educator came to me and he said, you know, by the time we get to physics in as a junior or senior in high school usually, the games over the students that you wanted to reach are no longer there. You can't help them. They've dropped out where the game is is mathematics. You got to come over to mathematics education and help us over here. And that was a huge turning point in my life was agreeing that I had to learn how to do mathematics education, because that's where I could have a greater effect on the lives of students I really wanted to reach. And what about you Thomas? So we came to graduate school thinking that it was going to focus mainly on physics education, having people learn physics. And I think there were a host of factors that coincided that substance will we change that trajectory. So I started graduate school in 2001. So a few weeks afterwards was was 9/11. And I think for many people in my generation that was also a moment of politicization as well. And in very soon afterwards there was the the war in Afghanistan, the war in Iraq. And so a lot of my early years in graduate school was very much influenced by the larger political context. So that's simultaneously a lot of my coursework ended up in ethnic studies and rhetoric and sociology thinking about ideology and post colonial theory. And the question that I started to really want to pursue is this question of how people make sense of issues of justice, how the nation moved towards the war in Iraq, even though there wasn't substantive evidence to back the effort. In many ways that's what I wanted to pursue. And yet I wasn't necessarily ready for it. And I don't think the field of the learning sciences was ready for those sort of questions yet. And so it's fascinating to think about how you grow with the field because I think those those questions are actually more acceptable now in the learning sciences to think about political change. It's really those processes of both the intellectual journey here at the university, but also the larger political context in substantive ways changed the focus of my research eventually to look at teachers, ideologies around race, racialization and racial justice. I love these origin stories because you know looking back and thinking about your work and what I know about the range of your work.
It makes so much sense how teaching and work at the graduate level really shapes the work that you do now. Both of you after your graduate career pursued different pathways. Jeremy, you went straight into a nonprofit research institute. Thomas, you decided to pursue a career in academia. And there are a lot of people who are at those crossroads right now. I'm curious about your decision making and what pulled you one way versus the other. So it's funny. You asked that question Monica because right after I graduated, I was convinced that I did not want to go into academia. So I started teaching high school again. As a graduate student, I don't think I saw some of the models in terms of how I wanted to engage in an academia. And just kind of the some of the cultural aspects of academia. And it was while I was teaching a few friends and few mentors told me about a position at UCLA. And UCLA as a department, they told me was a very, very different department. And it was listening to the podcast you had with Nicole and. And that as well, we're also describing just the culture of the department at UCLA at that time. And so given this, this story that I started hearing about UCLA. I went ahead and applied there and. And load me a whole once. When things worked out, it was actually really powerful to see that there's a different way of being an academia. A way that centers collectivity, a way that. Center's collective commitments. And that was a powerful, powerful learning that actually brought me back into academia. So you reminded that we can do this work differently. We can do it with a sense of kindness and care and commitment to each other and to the collective. It's a great story, Thomas. And it reminds me of the, the curviness of these paths, at least for me, it also wasn't a straight line either. I started out after graduate school. I worked for a little while. I had postdoc for a while in Australia. Then I worked for the, for for the Institute for Research on Learning, which was the spin off of Xerox Park. But it was being run by Stanford and Berkeley professors. So it's kind of on the boundaries of these different worlds. And then there came the conversation I referred to earlier with Jim Capit. And I just decided I want to work with Jim Capit. And so I did that as a non tenure track faculty member with the University of Massachusetts. That's where he was. And I used to call my office the University of Massachusetts San Francisco. And in fact, early journal of the learning science issues that have my name on it that has, that's my affiliation you mass San Francisco. I really had to make a decision for real at the time we were having our first child. Because I just felt the need for more stability. And that's when I really thought to myself what do I want. I felt like I really love interdisciplinary work. And there are university settings in which university work is valued. And really good initiative what disappoint work takes place. Thomas works at one of those with the AI Institute. However, I think also tons of stories where people do interdisciplinary work and the tenure committee sees that glass as half full. You did great work in psychology. What were you wasting half your time with the computer science department? You could have done twice as many experiments with college sophomores. That really worried me. And I felt like in the nonprofit world. There was no barriers to interdisciplinary work that was just the way work got done. And there was also a problem of status I felt to in the universities where many schools of education were looked down upon in their university hierarchies. I didn't like that. I just wanted to be a first rate citizen wherever I was. And so that made the university on appealing to me. I can some universities that wasn't true of, but it was a problem. I never minded writing proposals to earn my own salary and keep thing going. And so I thought, hey, I'll give it a shot at this nonprofit thing and see how it goes. That was the big downside at the time was, you know, you're going to have to find yourself. And that could seem a little scary, but I thought it was worth the risk. And you know, it's worked out one of the things I'm proudest about in my whole career is that I've never had to let someone go for lack of money. And I've run research groups up to about 100 people at one point. Never had to let go of someone go because we were at it money. So I think the nonprofit track it increasingly today, this is true. But the nonprofit track now can offer stable careers back to the beginning of the story. That's what I needed. Had to provide for my growing family. One of the reasons why I invited both of you here today is because both of you have done work around thinking about the role of technology and supporting learning. And I wanted us to think a little bit about the future of learning sciences as technology becomes more and more pervasive in our lives and inside classrooms. So let's start off this next segment of our conversation with this question. What's one word that you would use to describe your own relationship to technology? Think for me it'd be intentional. Say more about that Thomas. As a tool technology offers incredible possibilities. And yet if we don't close the exam and its assumptions, it can also reproduce hierarchies inequities and justices. And it brings upon us the responsibility of intentionality with which we use it. I was trying to come up with one word that captured two thoughts to contrary impulses. So I'm going to go like yin yang. And that that's really how I think of technology. There's always excitement and promise and glimpses into what could be. But there's also the realism that it doesn't come true that way. And sometimes technologies get out and they increase inequities, I should say. And so I think the awareness of that yin yang is really close to how I've lived with technology in my career. I wanted to ask a little bit about your current work and interests. Thomas, I think because of your teaching background as well, your work is pretty consistently highlighted the sort of the complex nature of the work of teaching and the work of learning. And you know, right now is the director of teacher education. I know that you're foregrounding that complexity and your work with pre service teachers. One of the things that you've cautioned the field about is these notions around replacement, right, technology replacing teachers in the classroom. And there's a lot of trepidation about that not only in education, but in industry at large, right. And you expand a little bit about why you think it's so important to center the deeply human work of learning in the face of all these emerging technologies. When it comes to educational technologies, I see two recurring themes that are to de center the teacher. And of course, like ed tech is diverse and I don't mean to make sweeping claims. It's important for us to consider some of these recurring patterns. So the first is about trying to control the teacher and minimize human error or variation. This is oftentimes referred to as like teacher proving the classroom. And so when it comes to say when film and TV first came out, there was the assumption that these technologies could record the best teachers and just make them available to everyone. And then the historical hindsight, we know how flawed this assumption was yet so many of new technologies are still based on the same underlying assumption. The second theme is perhaps more generous and tries to improve or perfect the teacher or give them more capabilities. And so here the assumption is students would learn more if teachers could be everywhere at once if they could know all the conversations that were happening at each table. And if they're able to support each student individually. So technology really becomes about helping teachers become omnipresent or omniscient in classroom. So there's two themes that have been recurring. But what we miss in this is that teaching and learning are complex and messy and our attempts to like tame or perfect them will be limited. But that's not necessarily a bad thing because it's it's the messiness that creates space for wonder and amazement and uncertainty and confusion and even helpful forms of frustration. So these are these are all experiences that are important to a more expansive view of learning. So at stake here what these views of technology is is our conception of our very conception of what teaching and learning mean and what it means to be human. So many of the discourses about the use of technology in classrooms tend to constrain learning to like humiliating knowledge or mastering particular skills or engaging in certain discipling practices. And you see this particularly in many of the very popular technologies that are are used in many schools now that tend to be based in more behaviorist models of teaching and learning. They're all essential to learning, but it's not everything. in classrooms or all.
also learning to become. And that's something I've been thinking about more and more, who are we becoming as we learn? So to be in dialogue, to listen and understand perspectives different than our own, to work with each other, to see our mutual humanity. And this is all deeply human relational work for which teachers are essential. And teachers again will use technology intentionally for multiple purposes. But the core work of teachers is irreplaceable. I love what you're saying there, Thomas. And it so resonates with the story I was thinking about, that one of the first questions here about, like, why did you choose this? I think I found a litmus test that's often useful as when you're talking to people. Do they think technology is really complex and interesting and rich and worth unpacking? And people are pretty simple. And what the people are going to do can be pretty underspecified, because it's not that important or the other way around. You were just talking so wonderfully about the richness of what teachers do. And I tell I really feel like technology, in my research, technology has served students and teachers best when it does a few simple, clear things. But the people part of it can't be reduced. It's really complicated. And so that's where my values live, is with the people who care about people and see them as really rich, complex, sometimes mysterious. But that's where inquiry has to lodge. It's with our view of understanding people. Jeremy, tell me a little bit about the kind of work that Digital Promised does. You're the executive director of the Learning Sciences Research. What is the way in which Digital Promise or DP-- how does that, as an organization, take up this research, this inquiry around the role of technology and supporting learning? First thing to understand about Digital Promise is the majority of the people are working with schools to implement programs or support schools in various ways. The Research Group is only about 30 people out of a total organization of about 200. So it's a smaller part. When I was doing research before Digital Promise, often researchers would have an idea and then we would get the permission of schools to work with them. But we didn't have much contact with school leadership other than getting permission. And to our surprise, nothing ever sustained, nothing ever lasted. Well, in retrospect, that's kind of obvious. Who makes things last? It's the school leaders who do that. So that's the biggest difference of Digital Promise. Its audience is a set of school leaders, for example, in the League of Innovative Schools. The League has about 150 school districts, and the members are a school superintendent. So I get the privilege of listening to school leaders and understanding what their problems are that they want research support to help with. And that's the biggest difference by far is our ability to listen. And hear things, of course, from teachers' perspective, but also from school leaders' perspective. And it takes some time. But we listen and we form a thing called the Challenge Map. And the Challenge Map is the most important problems that school leaders want help in solving. And we try our best then to conceptualize research projects that are grounded in what's becoming possible with technology and their skills as researchers. But that could help with problems in the Challenge Map. And we also proceed with the research in a way that schools are full partners in co-design and implementation. And if the research goes away tomorrow, they're going to keep implementing. Because this is valuable and important work to them. And so that's just a really different set of attitudes. And I think what Digital Promise has offered me is a structure that makes it easier to do that. You could do this when it just descrupt anywhere. But it's a lot for one person, one team, one research group to do. It's easier to do when you have an organization that's doing a bunch of that work anyway. And that really cares about those things. - I've been thinking a lot about the kind of opportunities and challenges that the field of learning sciences and computer-supported collaborative learning as well faces as these AI technologies become more pervasive both in our lives and then also in schools. I was thinking back to my days as a secondary student of doing research at the library looking stuff up on Encyclopedia Britannica and listed an alphabetical order. And using no cards to document my thinking and thinking about how to coordinate them and chronological order to reflect my argument. And the advent of the Google search engine, how that dramatically changed the kind of work that we do. Right? It shifted us away from thinking about acquiring information to thinking about evaluating them for their credibility and challenging us to think about how we synthesize so many sorts of information ourselves to really think about what our claims are, right? And so I think technology has this incredible ability to really shape us and how we think and learn. You know, in the last several months, chat GBT is on everybody's lips, it seems like. I'm curious from your perspective in what ways do anticipate technologies, these large language models like chat GBT? And what ways do you anticipate them changing how and what we learn? I can't ever forget the first time I read Larry Kupens' work because it was incredibly humbling, right? So as Larry Kupens writes, with every major technological innovation, there's both been exaggerated promises and fears about what would happen next. So this was true when radio came out, when film, TV, laser disks, all the other technologies that we've forgotten oftentimes. But many of the debates that seem brand new right now seem like they're recurring debates if we look back over across the century. So he brought up the example of search engines. And so in our lifetimes, we've heard similar arguments about things such as the graphing calculator, right? So whether a search engine is a graphing calculator, on one hand critics argue that these technologies would really prevent students from learning the fundamentals and would encourage different forms of cheating. And we see these arguments right now with chat GBT. But on the other hand, proponents argue that it would save students from the treasury of manual calculations or going back to your example of Monica going into the libraries. They're not how many of the listeners will remember the physical card catalogs that you had to sort through and find books. But if we could get past this sort of medial aspects of learning, then students can engage in high level thinking and big ideas. But despite these fears and promises, I think Cubans analysis holds true. Like schools really haven't changed much over the last century. So sure we have students typing up reports on Chromebooks or looking for sources on the internet as opposed to in a book or in Encyclopedia. But the fundamental way we do school hasn't changed. So from this perspective, I wonder whether a chat GBT will really change how and what we learn in the foreseeable future. For better or worse, it seems like schools have an incredible amount of institutional inertia that resists any form of change. But I think the questions that are coming up for me is how chat GBT can be understood in the larger political economy and what that means for schooling and learning. So what are the economic drivers of generative AI? What are some of the larger concerns that are rising in terms of copyright, especially for writers and musicians and artists and people who produce creative work? What does it mean for personal data? And also, like stepping back and forth, there how do we understand the emergence of these technologies in a time of increased political polarization and distrust, boss groups and parties? So I think the bigger question for me is how we understand where we are as a society, where we are as a nation. Going back to this question, what type of learning do we need for this moment? Which for me is learning that centers often to dialogue and recognizes our collective humanity? I agree with everything you said. I love it. And I wanted to drill into one subject matter, the one I study the most, mathematics, and just think a little bit about what could be different there. In mathematics education, getting students to master algebra is this enormous thing. And if you take the period of my career, for example, the amount of change we've produced in who learns algebra, how readily they learn it, how much pain is involved in that learning, we've barely moved to needle, really barely. And now we're at this moment. And we were there as Thomas points out before with graph and calculators, where, you know, the pain of learning the square root procedure, we don't need to do anymore. The pain of long division, not worth it. It's important to know how to graph, but once you know,
how let it device to your graphing for you, because potting point by point is not a good use of human minds. What's out human minds aren't that great at algebra, but computers can do algebra just fine. And could we be brave enough as a society to change the goalposts? And rather than penalize so many of our youth with this painful experience of this drudgery of getting through algebra, let's give them a tool that can do it for them. What would we do instead then? I think mathematics education has a great answer to that. It's the mathematics of modeling. How do we apply mathematics? How do we construct approximations of reality that are useful using functions, using symbols? Let the computer do some of the number crunching and some of the symbol pushing, but the art of modeling the world in a way that's responsive to our human judgment as to what is important to model, what are we solving for? Is the answer moving us towards the direction we want to go? Those are problems that intrinsically involve people as thinkers. And if I could wave a magic wand, my magic wand would make the symbol pushing part of algebraic, beat dramatically reduced, and give tools that can do it earlier, and refocus them on a different conception of what math is, that math is a tool for modeling your world. And for giving you power to make changes in your world. And I would spend a lot more attention on helping all students understand what it would mean to realize that power. And then they can power it by AI tools that can help them do some of the work. So that would be a fantasy of a really big change. And it would be really big. I mean, if you realize how much money, how much assessment, how much college admissions is all sitting on algebra one, if we could change that, it would be massive. I'm not that hopeful, by the way, about that, but I do think our society has to think about it at this point. And I believe some society out there, some country is gonna do this, and it's gonna be to their great advantage that they're gonna have a population that can leverage the math capabilities of AI models to do amazing things. And those countries that are still trying to get kids to push the symbols around, while they're gonna be falling behind, because we never get better. Which just, it doesn't seem like that's the best thing to focus human learning on. - Something I'm hearing across both of your responses reminds me of work by Gerard Fisher who talks about technology as a tool that can create sort of feasibility spaces, right? They can't do it themselves, but it opens up spaces for us to really rethink what it means to learn, how we learn, what constitutes the discipline. And I feel like both of you are saying things like chat GBT, there's always gonna be new technological innovations that come around, right? What it offers us every time these things, new technologies arise is the ability to kind of hold this mirror to ourselves in terms of thinking about what is the learning that's important? What is vital to our own human functioning to our understanding as our society as we move forward? Chat GBT and technologies that we gotta think about them as opportunities to rethink rather than tools that will automatically generate new possibilities for us. - What you just said is like my definition, that's how I live learning sciences, that the learning sciences is engaged with the idea that the future can be different for learners and technology can help us with that. But it's really also even more importantly engaged with how people learn and understanding that our school systems and our informal institutions as we've inherited them have done their best in their times, but they didn't understand learning as well as we understand it today. And they need to be redesigned as we've unpacked what it means to be a great learner. So technology is a tool for that unpacking. To me, that's the learning sciences is living with that set of possibilities and tensions. That's what it is to be in our field. - RSS makes me think about the chapter that you wrote with Barbara Means and Claudia Mezzarati in the International Handbook on Learning and Inquiry. And there you talk about sort of the importance of understanding the degree of ambition, right? Folks who are in the field, reformers are constantly asking for change on a large scale. I think I work in science education, right? And so we're constantly thinking about how do we transform learning environments in science classrooms? And one of the things that you say in that chapter is really trying to understand how degree of ambition shapes how we think about scale. And so my question is when you think about technology and its affordances and constraints, do you think about them as having like a small impact on a large population, which is something I talked about with Phil Vahey actually last season? Or do you think about them as having the ability to make more dramatic changes to our epistemological understandings around what it means to learn and what it means to teach? - I see technology as an infrastructure that it's not a direct causal agent of anything. It's a set of possibilities that we build systems of teaching and learning on top of. And that's super important 'cause it doesn't by itself do anything. And so some of the things that technology helps a lot with can be quite simple. I did a study with assistants, Neil Heffernin, Worcester Polytech, and homework support tool. And it was really helpful for low income students, our randomized controlled experiment found, really helpful for low income students to have homework support while they're trying to do math at home. What was doing, it wasn't rocket science. It was giving kids feedback on their answers and giving teachers insight into what students were doing when they were doing homework. But that small technological change with a system of teacher professional development on top of it did move in needle. On the other hand, we have examples like before a certain time statistics wasn't taught in mathematics class because, you know, you need too much resource to compute statistics. But as technology became available on calculators and whatnot, it actually became something kids can learn. And today data science, it's actually something that the Thomas has done things with can be a really exciting point of entry and growth area for kids in mathematics. All of that's only possible because computers were cheap enough and compute statistics. So there are some pretty huge changes that can occur with technology. But they're all based on like the statistics one, what kind of curriculum do we build? How do we work with math educators? So they can learn to add new subject matter to what they teach. It's all based on the systems we build on top. So that's my sort of both-and thing that yes, it can produce major change, but only if we're thoughtful about the whole system or trying to move and technology is just a piece of that system. Thomas, you've been producing a lot of work in the learning sciences that asks us to think about the ideological stances that underlie our efforts to support and improve learning. Let's think a little bit together about Chagy PT. And I think this is linked to something that you've said before. What do you see as the ideological stance that's sort of underlying its creation and sort of rapid uptake? And from your perspective, what's sort of like the best possible or maybe the least ideal outcome of these technologies inside classrooms with teachers and students? Still thinking about the comment you made earlier about technology as being a mirror and the role that it plays for us to then think about what learning should look like and Jeremy's reminder that the role of the learning scientists in this work, understanding learning and deeply understanding learning and from this vantage point, which we hope the world's gonna look like. And I think to do this, it really comes back to the question that you're asking right now and that is also understanding the ideological assumptions that underlie so much of these technologies. So the question about the ideological stances that led to the creation of technologies like Chagy PT is really fascinating and light of Jeffrey Henson's very public resignation from Google last month. Hinton was a cognitive psychologist, a computer scientist. He often gets those nomenclature such as being the Godfather of AI or the Godfather of deep learning. But he's made a splash in the media when he resigned from Google because of his concerns about the risks that AI poses. So when he was interviewed, he expressed regret for this work, but justified it with some very common stem tropes. So the first was the justification that it was a compelling and challenging problem and that as a scientist, he just wanted to solve it. The second was if he hadn't done it, someone else would have. So I think these are some of those ideological assumptions that underlie a lot of technology development. So it's more a question of can it be done and emphasizing that can it be done versus should it be done.
And I think that's where the ethical questions around technology and technology development really come into play. So his resignation, these cautions and reminders are, is a really deep reminder that new AI technologies have implications again for who we're becoming as human beings. And as an aside, I also just want to acknowledge that a number of others have been expressing these concerns for years without the same media attention. And so it's important for us to recognize a longer history of this critique. But these questions can't be narrowly defined as technical problems that only require STEM expertise. Our collective decisions also need to be guided by the humanities and by the social sciences. So I realize a much of what I've said today is rather skeptical of the large scale use of technologies. And again, I want to distinguish the thoughtful ways in which a number of learning scientists approach the use of technology and the way it's taken up at scale oftentimes in schools. But I see this as a healthy skepticism that we need. And acknowledging that these technologies have incredible potential to support us in analyzing, representing and communicating data. So as a tool, they're the potential to extend human creativity and capacity. So I can be an ideal situation. I hope that these technologies enhance teacher judgment and support relationships between students. One of the hopes is that we can design an AI tool that would represent relational dynamics in the classroom, back to students and teachers so that it could support the reflection and their intentional actions toward more collaborative relationships. There's much to be figured out here. And there are a number of ethical questions that are emerging along the way. But I think it hints to the types of possibilities these tools or new AI technologies might have in classrooms and schools to enhance human judgment, enhance human relationality as opposed to minimizing it. Bill F. Nuitama said, "One of my calls to action for learning scientists would be to get more engaged with policy." For a lot of time, I don't think we had to because nothing we were doing was having such a huge impact. But we're in a very odd state right now. You can flood schools with something like chat GPT, but if you try to put a new swing on a playground, you have to cough through tons of regulation, procurement regulation, installation, regulation, red tape. Environmental impact statements, what not. We too regulate things kids use in school, but as a society, we haven't regulated technology much at all. And I don't know if that can last given the disruptiveness and the power of these technologies. But my fear on the other side is when you say that you could get some really silly, stupid, awful ways to regulate this too, I think learning sciences could help with the core of it. We value how people learn. And we want to think about allowing schools to use those things that are going to help the learning process and help the teaching process. And being careful about those things that may have unintended consequences, which often occurs in our research. We think something's going to be great and something pops up that makes it not great. So I think it's really time for us to add a new dimension to what the learning sciences is. And that's a dimension which is engaged with shaping policy. In 2017, Thomas, you published a paper in the Journal of Learning Sciences with Ayush Gupta and Andy Elby and Chandra Turpin. And you wrote at some level, all learning is ideological, setting theory. And by valuing certain domains of learning and under valuing others, communities and societies, I prioritize certain ways of making sense of the world. And if we do not critically examine the ideological nature of the settings in which learning takes place or the ideological processes of learning, we risk reproducing and creating new forms of inequalities and injustices. Thomas, I'm curious, what would it look like, feel like our sound like for learning scientists to really examine these ideologies that underlie both the development of AI technologies and the days to come? And what kind of sociotechnical infrastructures can you imagine coming into view if we do this work in an ethical and human way? A question that's been sitting with me is when we learn and especially when we learn with technologies, could we become? And I think that's been a recurring theme in much of my work. So when I think about the article that you mentioned with Ayush and Andy and Chandra, a takeaway that struck me is that most of these undergraduate engineering students who were learning about ethics of militarized drones, they weren't just learning about drones. They were also learning through their discussions to devalue brown-bodied civilians in the West and South Asia. And ever since we worked on that article together, the question of who we are becoming as we learn has become frontman central for me. So it's a parallel question to when I worked with Maria and Janet on our article on racial data literacy where we tried to show the students were learning about race just as much as they were learning about data science. Again, ostensibly, they were just learning about data visualizations of movie-virtual patterns, but it became a deep contestation over the media race. So what would it feel look, sound like for learning scientists to examine the ideologies that underlie the development and use of AI technologies? Again, so much of it for me comes down back to questions of how we conceptualize learning, who are we becoming through this form of learning, to what ends are we learning, and we have to ask these questions in light of educational technologies that we design if we are to leverage the strengths and the analytical tools and the theoretical tools from the learning sciences to shape technology in a way that's going to move our society forward in ways that really center equity and justice. In Jeremy in your review of the International Handbook of Computer Supported Collaborative Learning, you invite current and future scholars in the learning sciences to lean into disequilibria, sort of the unresolved tensions that can lead to transformations in our field. What kinds of unresolved tensions do you see in how we conceive of and/or use AI technologies? And what do you hope to see coming down the pipeline in our field? Well, that seems like a big question. What I was writing about when I wrote for that International Handbook, which is a great volume, was really the creative tensions in the work. For example, several chapters had people doing machine learning coming into CSCL, which had been incorporating these technologies over time, but more and more was happening. So one of the disequilibrium was created by a new technology. But there are also tensions about among different levels of analysis, places where the work is occurring, thoughts about what the goals of collaborative learning are, including a shift towards metacognition as an important goal of participating in collaborative learning. And so I think in all science, actually another famous Thomas I learned from in my undergraduate days was Thomas Koon. And we go through these periods of stable science and revolutionary science. And it's exciting when things are being shaken up and new possibilities are emerging. One of those are fields coming together or techniques coming together with new fields. And I think we're out profoundly at one of those moments right now for the learning sciences. And there is a kind of normal science learning sciences that is a stable learning sciences, which could carry on doing the same kind of investigations it's done for a very long time. But I think if we don't pay attention to disequilibrium, it would be a more and more irrelevant learning sciences. I think the relevant learning sciences of today is going to have to engage across these new technological possibilities. It's going to have to bring together new forms of expertise. It's going to have to ask different questions like I grew up in the learning sciences asking how can we learn math better and listen to the difference between that and the kinds of questions Thomas has been posing. So I think the healthy future of this enterprise of the learning sciences just really depends on people walking, getting right up to the edge of these disequilibrium playing with them. This has been such a fascinating conversation. Thank you both so much for all of these insights. I've been waiting to talk with the both of you and admiring your work truly for a long time. Thank you for this opportunity to chat. Thanks a lot right now. All right. Bye. I'd love to hear what you took away from this conversation and connections that you see to your own work. Send us an email at
[email protected] and find us on Twitter at humans.lspod. This podcast is co-produced by Andrew Kersak and
and Monlin Monica Co. Our work is made possible by the Learning Sciences Research Institute at the University of Illinois at Chicago. Thank you for listening. (upbeat music)
PodcastTranscript AI