Rebecca Winthrop | Rethinking the Purpose of Education in the Age of AI | Talks at Google

It's my pleasure to welcome Rebecca Winthrop here, who's become a friend over the last year and a half. I think we've been meeting regularly. And she's the director of the Center for Universal Education at Brookings, which is dedicated to making sure that every child can thrive through education. And she leads cutting-edge research and initiatives aimed at transforming education initiatives around the world to support children's learning and development. Her expertise is sought by many people, including the UN, the White House, Fortune 500 companies, but also parent networks and teachers and school districts and so on. Her work is centered on developing and advocating for evidence-based strategies that bring people together all of the different factors in the education system, the families, the educators, the policymakers, and companies that are creating products for this environment to help children maximize their potential. She holds a PhD from Columbia University Teachers College, an MA from its School of International and Public Affairs, and a BA from Swarthmore College. Welcome, Rebecca.

Thank you. Lovely to be here. So—

So last year you came out with this book, The Disengaged Teen, and it was a blockbuster from what I can tell. So many people I know have read it. And core to the book were two dimensions and four learning modes. Could you walk us through the framework you created? Because it was based on a lot of data, and I personally found the framework extremely useful in interacting with students I see in schools. And I know other parents tell me about how useful that framework is. So could you draw that framework for those who may not be familiar with it?

Yes, yes, absolutely. So my co-author and I, Jenny Anderson, we were really interested in a very simple question, which is why don't kids like school? They love to learn, kids are hardwired to learn, but we know that in 3rd grade, 75% of kids say they love school, and by 10th grade, it's flipped, 25%, which is opposite of adolescent development and brain development. Kids are like really on fire in adolescence to learn and explore. And so what we found out that was that it isn't so simple as either a student is engaged or disengaged. They show up in these 4 modes that you reference. So it's passenger, achiever, resistor, explorer. And they kind of are what they sound like. So kids in passenger mode are the kids who are coasting, doing the bare minimum to get by. They might really like going to school to see their friends or for, you know, debate team or sports. Class team participation, but they're really not that interested in what they're learning. Interestingly, we found that a lot of kids had really good grades in passenger mode. They just weren't challenged. They were pretty bored. And the number one thing that kids who were— spend a lot of time in passenger mode told us when we said, tell, you know, tell us about school, is they say, oh, it's boring. Like, that's it. It's boring. I don't see the point. And they were really sort of missing why you have to learn something in the first place and its connection to how you use it in real life, in the world, jobs, whatever it may be. Apparently every kid in America hates the Pythagorean theorem. Maybe you guys, I think this might be the one place that doesn't hate the Pythagorean theorem. That could be your new like HR hiring test. Did you hate the Pythagorean theorem as a child? If not, you're hired at Google. Yeah, but that was the thing that was quoted to us so frequently. Why do I have to learn the Pythagorean theorem? What is it good for anyways? So that sort of ability of schools and educators and even parents to show why what you're learning in school is important, it is needed, and how it's used and how it's helpful is really what characterizes kids in passenger mode. And we found that about 50% of kids, middle school and high school kids, said that they regularly spent time in passenger mode in school in the US. Then, so that's sort of disengaged and pretty passive, not a lot of agency, not a lot of engagement, though they're physically present. Then you've got achiever mode. These are the kids that everybody loves. We probably have a lot of achiever modes here too. Kids who spent time in achiever mode, trying to knock it out of the park and get a gold star on everything put in front of them. And in many ways, these kids are developing really good skills. They're goal-oriented, they're organized, they're ambitious. That is all positive. The issue we found is that the number one thing we heard from kids who, who spend a lot of time in Achiever Mode in about the thing that they hated the most in school was when their teacher didn't tell them exactly what to do to get an A. That just pissed them off. Super frustrating, hated that teacher. One of my students, how am I supposed to get an A if you don't tell me exactly what to do? That's so unfair, you know. And we did find actually that it was pretty easy to slip from happy achiever into unhappy achiever mode where kids in achiever mode, they're just really indexing on the outcome, not the learning process, not necessarily what they're learning, not necessarily their connection to their interests, not even necessarily how it applies to the real world. They're just like, I know I need to get all A's to go to a good college and I'm gonna do it by gum and I'm gonna, you know, do my best. And a lot of those kids It's very— felt like they had to be perfect. And it's impossible to be perfect, and there is absolutely rising rates of social perfectionism, which is where you feel other people want you to be perfect. Parents, unfortunately, have had a big role in this. And because they felt like they had to be perfect, they would really melt down when they would get a B+ or a B-. It was sort of the end of the world. Not resilient, and a lot of mental health problems, actually. Really high rates of anxiety and depression. So that is sort of the dark side of Achiever Mode, in addition to the, I need to be told what to do because that's how I'm oriented. That's what I think school and learning is about. So those young people are really engaged. They're physically present. Care about school, they're trying really hard, cognitive engagement, but they're not really proactive around their initiating a lot of learning. They're becoming great followers, great followers. And then you've got kids in resistor mode, which is like the opposite of the kids in achiever mode who everybody loves. The kids in resistor mode, everybody's tearing their hair out. These are the kids who are avoiding their learning, not turning in their homework, skipping class or being very disruptive, talking in class or having a fight out with a coach or whatever it may be. It's really hard to parent these kids, really hard to teach these kids. And so when you think of a disengaged student, people think of kids in resistor mode. But what we found is that they actually have a lot of agency. They have some gumption, they have some chutzpah. They just are pointed away from their learning. They are actually exerting initiative to get out of school. And they're often very creative in how they try to do that. And if you can pivot the learning environment, it really is all about the learning environment that makes kids get into these contexts. If you can pivot the learning environment in a way that they get excited about something, they move much more quickly to the last mode, which is explorer mode, much more quickly than kids in achiever mode, for example. And so kids in explorer mode is really where you want kids to spend a big chunk of their time. They don't need to be there 24/7 because these are the kids who are really interested in the learning process and they are taking initiative in some, could be small ways, could be big ways, to make the learning environment more interesting and supportive to themselves. So they may be asking a teacher, hey, could I write my paper on this topic? Or hey, could I do my project with a friend? I think actually if I go on YouTube and do a deep dive, that would really help me. Can I do that? These are some small ways you can really take initiative. They're very interested in what they're learning, and they're not necessarily indexed on just the outcome. And it turns out that when you're in explorer mode, you get the best grades in your school. There have been within-school comparisons, class classroom to classroom when kids get the chance to be in explorer mode, they beat out academically the kids in achiever mode. And one of the ways you can find out if your kid is in explorer mode is they come home from school and you say, "Hey sweetie, how was school?" And instead of like, "Fine, boring, okay, good," you know, not much else. "Mom, do you know what I learned in school today? Did you know there was a tornado in this blah, blah, blah, blah, blah?" And then just whatever it is, whatever it is, talk about tornadoes, they don't stop stop talking about tornadoes. They can be a little annoying, but like they're on it about tornadoes. They do deep dive on YouTube about tornadoes. They talk about tornadoes at dinner. So they are in it, and that is a real sign that they're super engaged in the learning journey.

So do you think of this— so it seems like the whole system, the parents, the teachers, everybody plays a role in, to some extent, what mode the kid gets into. But also it feels Just looking back on my own childhood, right, in India you succeeded by being in the achiever mode. Achiever mode, yes. But there were other things my school had, like a chemistry club, right, where you could be in explorer mode. So do you see that dichotomy happening with some kids where outside of the things where they are heavily measured on, they actually get into explorer mode? Or is it they have to do it in school?

Yes. No, you absolutely do not have have to be in explorer mode in school. We found a lot of kids in passenger mode in school and a lot of kids in explorer mode outside of school. And it really doesn't matter where. What you want is kids to have explorer moments that are building those learning-to-learn muscles, the sort of ability to persist and pick yourself back up if it gets hard, to dive deep, to self-direct, to find a way to make your learning journey better for yourself, 'cause you may know where you need help and what you need. And outside of school is just fine. The part that I worry a lot about is how little time kids spend in explore mode in school. I'm not saying it needs to be 100%, but we found that less than 4% of kids said that they regularly spend time in explore mode in middle school and high school. And that's really just too small.

And you think it's tied to the stakes?

Yes.

And so you need lower stakes to get kids into the—

I think it's tied to the entire design of schooling, which, I mean, basically, I work around the globe. Basically every school system is designed the same way at its core. And the core is, ranking and sorting children into higher or lower. Why else do you have grades and pass, et cetera? Largely, that is an artifact from when school systems— this idea of public school systems, universal education is a radical idea that's been around, born in the mid-1700s just as an idea. And it really came out of this idea that— or grew, I would say, scale— of we need to find a feeder for higher education. Because higher education, this idea that truth is not just in the religious institutions, truth is accessible to the general— anybody in the population is a radical idea that sort of, again, the Western universities a couple hundred years ago, 1600s probably, and then they needed feeders to go into that. And as much as we've tinkered around the edges, people talk about sort of, oh, it's an industrialized model, and certainly sort of this idea of Taylorism and batching and bells is an artifact, I would say, of sort of a manufacturing culture. But school systems in industrial, spread in industrial countries and non-industrial countries. So it really is the sort of pipeline for higher education and a feeling of you need that to grow your economy and be part of the modern world.

Wonderful as higher ed is, it is a selective system. And so the sorting plays a role there and for jobs later on. And that's even more true in developing countries where there are fewer jobs, fewer positions. And just a small fraction get through. Yeah. So in all of this, the people really matter. And so in our work at Google, we are focused on human interactions in the learning education experience. The magic of the teacher and what role does that play and how can we enhance that, how can we help them, especially with school learners in high schools and middle schools. You've often argued that the relationship between schools and families is stuck in compliance, in what has to happen. And the question is, how do we move from that compliance way of thinking to something that's less about monitoring students to more viewing them as partners, as motivating them intellectually to get to that explorer mode? What are the things that unlock that?

Right. So one of the things we found, because in the book, the sort of second half of the book is a bunch of strategies of what can educators do, what can parents do, what could school leaders do. And there's a— they're all sort of packaged up in a toolkit for the different modes because you kind of have to do slightly different things for each mode. So if you have a kid in passenger mode, one of the classic dynamics that is going on— I don't know if anyone would be familiar with this, is the sort of nagging procrastination loop from hell where the school is like, "Your child didn't turn in their homework," and the parent's like, "You didn't do your homework," and then every day the kid comes home is, "Did you do your homework? Did you do your homework? Did you do your homework? Is your homework done?" And it turns out that nagging shuts down the problem-solving solving part of adolescents' brains. This is actually one of my favorite studies in the book and my sons, my kids will attest. I have one here today that you can grill them afterwards.

Is it true of management too?

I actually wonder about that. I actually do wonder about that. I think there might be a whole parallel of explorer, achiever and what you have to do to empower people in adult life. But I don't know, 'cause I didn't research that. But the study is, it's done by neuroscientists. They run adolescents through brain scanning machines, and they're listening 3 different times. They listen to a recording of, it's always the mother, the mother's voice. And, you know, one time it's something you know, normal, like, you know, "What do you want for dinner?" or something. Another is positive, like, "Hey, great job on your, you know, swim meet," whatever it is. And the third one is a nagging recording of the mom. "One thing that bothers me about you is..." And then they fill it in based on the kid. And when the kid hears that, the problem-solving part of the brain shuts down. So the more you nag, the more the child can't— is gonna procrastinate and can't even quite get it in gear anyways. And there is a lot of, again, neuroscience, we were talking about hidden brain before, like scientific evidence to show that procrastination is often an emotional reaction to sort of a negative feeling of you can't get something done. And so I'll just, let me just do a little more Zelda, you know, or whatever it is, whatever they're doing, scrolling on TikTok usually. And so, You do have to switch the dynamic. So on a parent side, the, you know, one very practical strategy is to convert from sort of nagging procrastination loop to a, we call it a learning to learn loop in the book, of helping kids make a plan for what their homework is gonna be, help them, let them execute, and then give a little time to reflect on the plan, maybe the next day. See if it actually worked, because that actually builds metacognitive muscles to be able to learn on their own. Because what homework is, it's important to practice and sort of embed knowledge in you, so that's why teachers assign homework. But really, the point is not to get the homework done that day, it's to help kids develop independent learning skills so they can go on and learn for the rest of their life. We were really shocked at how many kids actually don't know how to study, 'cause schools teach you what to learn. They don't really teach you how to learn. So if you have a kid who struggles with this, it really is like make a plan. And sometimes it's like, well, how do I do that? You know, 3 things. What do you have to do? How long will it take? And when are you gonna do it? Like work on that for a month. Like, again, don't expect, don't expect this in, you know, 5 minutes. And, you know, is it, are you gonna do your math before dinner or after dinner? And how long do you think that problem set is gonna take? Just a little bit of modeling. And then when they get that, you will have to let kids, again, this is kids in passenger mode, try, do it on their own. And if they bomb and fail, they will learn. And we're not advocating like, let your kid fail out of freshman history, bomb the SAT, not turn in their college applications, not high-stakes stuff like that. But there's a lot of time during the school year when they can practice on their own.

So the ability to— again, it comes back to the stakes being low enough that you can actually do the learning thing. Yes. Because the learning thing has to permit failure.

Yes. Because that is what learning is.

Yeah.

Yeah. At the core of learning is doing something wrong. And you go, oh, OK, I did. I have to adjust and pivot.

Yeah, that's what— it's true for machines as well as for humans.

Yes.

So last year you joined us as part of our AI for Learning Forum. You had this great talk about the purpose of education, which I thought was really wonderful. And you're currently leading Brookings' Global Task Force on AI and Education, which recently released a new report, A New Direction for Students in an AI World: Prosper, Prepare, Protect. Many of the findings of the report are reflected in our own position paper. Both the optimism as well as the cautionary aspect of it, right? And as we look into the future, like, what are the few— what are the things, the risk callouts that you want to— that in the report to draw that we should be thinking about? And how do you feel like companies should be—

Tech.

—trying to— tech should be trying to address those risks as well as realize— and realize the promise, because there are the two halves of it. Yeah.

I'm so glad you asked this. You have a captive audience. This is so exciting. So we did a big task force at Brookings, which I led. We led it the last 18 months. And it was really just a year of sort of deep dive research with the main question was, are we on the right track when it comes to generative AI deployment and students? And we were looking at primary and secondary students, K-12 in the US. And we were looking at— we were really focused on students and their learning and their development. And the reason we did it was because we don't want to see the same movie we saw with social media where we didn't pay attention. We, people who care about kids and learning and their development and their growth, didn't— weren't really at the table, and we frankly were a little asleep at the wheel. I think as a sector I can say that in education. And it was like, oh, well, that's that thing out there. It doesn't— what is it? We don't know. We don't care. It doesn't affect us. Learn your Pythagorean theorem. So let's not— Stay bored. Right? Let's not make the same mistake twice. And so we called it a pre-mortem because it's very early days. So we shouldn't necessarily wait for all the, you know, 5-year longitudinal randomized controlled trials to come in. Let's do a quick look and start a conversation. So that was our goal of like, you know, where do we think we're headed? And really what we found is that currently the risks are overshadowing the benefits as GenAI is rolling out. And it's largely because It's not because there aren't benefits. There are benefits, but the benefits are from very narrow strategic GenAI deployment with vetted content. So kids are actually learning real things rather than weird hallucinations. And this part is even more important, embedded in a really good pedagogical experience so that it's actually helping kids learn. So great, great, you know, use cases for teachers saving time and being able to level Learning materials, really helpful for neurodivergent kids. One of the most sort of, you know, moving cases I always like to cite is for kids with aphasia who have communication problems. They're now being able to develop synthetic copies of their voice and they can talk to their teachers and their peers in the classroom for the first time. I mean, really game-changing stuff. I think assessment can really help, could be amazing, really help see the learning journey across a wide range of skills. So there's really great use cases. The problem is that narrow AI use and deployment is not really how most kids are experiencing it. Most kids are experiencing GenAI through AI companions or frontier model like you guys, Gemini, and all the many others. Most kids we, in our study, were using ChatGPT. Unscaffolded, long, open-ended, direct conversations with frontier model chatbots that are really not made for learning. And it doesn't matter if you have ChatGPT Edu, right? Because right next to it is the give me the answer version. Kids are going to pick the give me the answer version because they're not lazy, they're not bad. That's just how human beings are. Give me the answer, why would I take all this time? You have to, it's the explorer mode kids who would go through and really learn, and that is the far and few between. So that is where the risks come from, and the risks are bad, you guys, they're not good. I mean, it's really around, I'm starting to call it cognitive stunting. The word everybody uses is cognitive offloading. So it's not just the kids are skipping steps and doing their homework and just asking, you know, their AI chatbot to do it. It's that is missing the opportunity to practice and build learning skills and embed content skills. And we know that if you, even if you put a couple sentences in an AI chatbot and then it writes the essay, kids don't retain this. I mean, teachers told us over and over, oh, I brought a kid in, they had an essay, I said, oh, tell me about your argument. They couldn't remember what they wrote, even though they said, well, I just, I gave it my ideas. It's not the same. So I'm really worried about hurting cognitive development, worried about social-emotional development with AI companions. As you all, I'm sure, well know, now 1 out of 3 kids say they prefer talking to an AI friend friend or companion equally or more than a human being. This is US data from Common Sense Media. And even kids would talk to us and say, you know, we're developing different emotional muscles. And again, the sycophantic nature of AI companions, never mind the safety and security stuff, which is really, really problematic, just in terms of learning. If you're used to always being agreed with and then you come into a classroom and the teacher's like, that is incorrect, you did that wrong. Like, you have to be able to take feedback. Like, so that, you know, what is that going to do for kids' ability to learn, take feedback, grow? We also found a really worrisome sort of degradation of trust in the teaching and learning relationship between— it's not just that teachers are not able to trust, feel they can't trust, and I think legitimately, a chunk of their students' work. Like, I don't know if that's their real work. And, you know, there's AI humanizers, and then there's the fake typers who, even if a teacher is watching in the Google Doc, there's another, you can put it in, you know, Gemini or ChatGPT and then put it in a typer, and then it goes into the Google Doc and looks like the kid is typing it. So there's— you're never going to outsmart the fox. You know, like, you're never going to— kids will always find a way. People will always find a way. And so teachers are saying, I don't know if that's my student's work. But students are also saying, oh, that was an AI assignment. I don't think my teacher is grading my work. I think they ran it through AI. And that's really interpreted as a lack of care. And then you've got parents. We've heard from teachers who are really questioning educators who get assignments back graded from their students, run it through an AI chatbot, grade this, it comes out in a different grade, and they go back to the teacher, hey, you misgraded it. You know, Gemini or ChatGPT told me differently. So it— that's problematic. And you don't miss trust until it's gone. Because teaching and learning is built on trust, and it's quite worrisome to me about what that means for the future of, A, developing kids' individual capacities in a teaching-learning relationship, but also schools are one of the first public institutions to introduce kids to people outside of their immediate family and their neighborhood. It's how we develop basic civic awareness of how— and learning to live together with other people and get along. It's— and if you have a really cynical vision of what school is all about, I do worry long-term, what does that mean for democracy and our civic health?

The flip side of this, though, is that The current system is far from ideal. Indeed. It's failing a lot of people. And around the world, it fails people in dramatic ways. I mean, the people you're seeing, talking to here, are the people who succeeded. But for a lot of people, there are things they didn't understand, nobody helped them understand. And now for the first time, there's something they can go to to help them with that gap. And when we talk to students, we do see that as well. Yes. And I see that in my own life. Right? There are things that I didn't know. I didn't have the patience. I also didn't know how to ask the right questions in areas I was interested in. In but didn't have an expert to talk to. So there is that other side of this that is this huge possibility of personalized, individualized ability to help me. Yeah. Where I am broken, where— which is not possible on the technical side of things. While the teacher still has to play this role of inspiring and like guiding and the relationship, I fully agree with the key importance of the teacher. How do you see that part of it integrating, right? Because it feels to me that a single human being can't do that with 40 kids. In India, it was sometimes 60 or 100 kids when I was growing up. And so there's something tremendous that's possible on the other side, right? So how do you see that manifesting?

So I think we really saw it in two ways. So one is the individual kid who's in explore mode, and you do talk to those few kids. Who are in Explorer mode, and they are doing amazing things. Like, talk to a student who's like, oh, I took all my notes from science, I put in and LearnLM, I made a podcast, I made flashcards, I made, you know, and she's like got her whole like incredible per— it's like a personalized, you know, learning coach right at her side. And she goes, walks the dog, listens to her podcast 3 times, and then she takes her science test and it's great. Like, she's doing all sorts of cool stuff or in Explorer mode and really into a specific issue, like, "I really wanna learn robotics, this is amazing, I'm super excited," and just like can go incredibly deep. So that's happening despite school, right? That's kids on their own accessing stuff. I think in a more sort of structured way, again, one of the benefits we did see in Callout was, to your point, this idea of personalizing learning. So any— again, the vetted content, the digitalized textbook that has AI embedded, and a kid is reading a passage 2 or 3 times, it's like, I just— I don't understand this, can you explain it to me a different way? Really impactful, really helpful, especially for multilingual learners, kids with learning differences. Like, that is really helpful. The question in my mind that I don't know how to Mary, is how do we just do that without all the other sort of sprinkle the frontier model chatbots around to everybody? So, I mean, that to me seems like a really joint big quest. I'd be curious how you guys are handling that. But I agree with you, that is a real potential benefit.

Yeah, and actually we see it also, like, as I was saying, when we talk to students, we see it as a real benefit in many of the students. And it varies by group to some extent. But yeah, you also, this is a minor point, but you labeled this Prosper, Prepare, Protect in that order. And I was like, huh, that's interesting, an interesting choice of ordering. Why did you choose that choice?

So in this, AI Task Force report, we have basically— it's actually 12 recommendations, but we put them in 3 big categories: Prosper, Prepare, Protect. And each one has sort of 4 specific recommendations. And we had big debates about the order. And really what they are— Prosper is what can we do to shift the teaching and learning experience so kids can prosper in an age of AI. Like, where, where is— should pedagogy be AI aware? So that means, you know, if you are giving an assignment that kids can just go home and stick it into a chatbot, don't give the assignment. Shift it up. That's AI aware pedagogy. AI assisted pedagogy, which is what we were just talking about. How do I help my kids who are, you know, teachers have multiple learning levels in each class they're teaching, it's very difficult. So, you know, how can I use AI to help all kids get, you know, access the content, that sort of AI-assisted? But then there's also AI-resistant. What are the things that young people need to do without AI at all? And it might be without technology at all. Like, you know, let's make sure kids are, especially with, you know, the amount of technology I'm talking about communities that have access to it, in their lives. Let's make sure schools carve out time for kids to look at each other eye to eye and have conversations and work together and get outside and move around and be embodied. Kids need that desperately for their development. We will not evolve as a species in the next 10 years or even the next 100, right? It took a long time for us. To evolve as human beings. We need relationships. We need in-person embodied experiences. So that would be the sort of AI-resistant pedagogy. So that's Prosper. That's what can we do within education to shift our teaching and learning. And we put it first. We had lots of debates. Should we do Prosper, Prepare, Protect? Protect is really holistic AI literacy. I do think it's time that we Everybody should learn computer science. We all learn biology to know how the living world works. Kids spend a lot of time online. Everyone should learn computer science, how computers work. You know, that's what we mean by holistic AI literacy, like really understanding how the online world works and how you can use it to advance your ends, not just how to prompt engineer. And then the protect is the obvious. Let's make sure that there is children are safe when they use these products because many times they are not. Some of your competitors have terrible, terrible track records with child safety. So we originally had it Protect, Prepare, Prosper. And then I think we had a big debate saying, you know, we don't want people to just be stuck on the Protect. It is a, it is essential. It is, it is, it's exactly, it's necessary but not sufficient. What we want is to have an education system that is better. Yeah. Where all kids thrive and where AI can be really used in a very strategic, helpful way to help all kids learn and grow. That, and that's the sort of Prosper vision. So let's put that up front.

And some of it requires creativity in the teaching system with teachers because teachers do such an individualized thing. They are relating teaching to students with knowledge, and that's an individualized process. So they need to be creative with these new tools that have come along, right, that changes the system. And I think I do feel that that's central to the success of this. How do we serve teachers, really?

Absolutely. And I think that might be the biggest thing that we should do, the biggest lever, is how do we support educators? Educators. We've talked about this a lot. I'm like, I want to do a next book on how do you support educators to be in explore mode, because I think In an age of AI, they have to be. Yeah, yeah. The world is changing. There's no other way. There's no other way for them to really help their kids prosper, so to speak.

So you also often use the framework of purpose, pedagogy, and position. And you had a very nice talk about the purpose of education and how it's been seen through the ages. And the purpose is often like the neglected thing in the context of reform. We assume a purpose and then think about reform in that context. How do you think society can come to consensus on what the purpose of schools is, which some of it is sorting, there's no getting around that in many contexts, but some of it has to go beyond that. So how do you think about that?

So one of— excuse me— one of the things that we have done again at the Center with colleagues is think about how do systems transform. Not just sort of incremental shifts, but how do big systems transform? And the, and the system dynamics field, which exists, has a lot of interesting frameworks. But the core is that there are these sort of hidden dimensions of systems that we often don't pay attention to, and they are mindsets mental models, worldviews. So a very easy example— well, easy for me because I'm not in the sector— is the US health system is really focused on— the purpose is to treat illness. It is not to promote wellness. That's why there's no— you can't get preventative medicine, you know, in doctor visits paid for. If you break your leg, you can get that paid for, for your insurance It's sort of treating illness. So we really looked at those— at the levers of how you transform systems. And one of the main things is to tackle these hidden mindsets, which shows up in what everybody thinks the purpose of a system is. Is it to treat illness, or is it to promote wellness? And so for education, we know that you need a shared purpose. It's not enough for a superintendent or a state secretary to sit and develop their vision. In the U.S., you have a lot of people working on this through their Portraits of a Graduate approaches where they're going out asking parents, asking teachers, asking students, asking, you know, employers, asking legislators what should a What would it look like in this state, could be a district, but usually in the state, for a student to graduate and we say they're successful? Like, what are the competencies, what are the mindsets, what are the characteristics? That's a great example of trying to develop a shared vision of what a system's purpose is that we often neglect because the purpose piece Once you get that right, the other pieces of reform, which are still needed, often fall into place. And they are things like shifting power dynamics. So who's at the table to define and give feedback on is your system functioning well? Are parents invited in to give their two cents? Are students? Or shifting measurement. How do we measure things? Or how do we, how do resources flow? And often what we do is we start with those. Let's shift how we, you know, shift resource allocation, but not shift the purpose vision. And so you kind of get stuck tinkering around the edges. It helps-ish, or it doesn't help, it depends. But you don't really make a big transformation. So we do need to spend time, especially with AI, which is a task task completion technology, and much of education systems are task completion assignments. And so we really need to think through what we do with schools and what schools are for and what we want kids to come out with on the other end.

Yeah, I think that's a— that's a— the central question today, and how— whether— how that change happens, whether it happens, where it happens. And, and are there some school districts— like, America is an interesting place because so much independence in how school districts and all work, and one of the places where you can imagine a lot of innovation happening because of that possibility to vary. So thank you so much, Rebecca. Of course. It was such a pleasure talking to you.

Pleasure. Pleasure.