Portrait of an Educator in the Agentic AI Era

In this episode of Start With Questions, Kristy Lundstrom is joined by Vriti Saraf, co-founder and CEO of Ed3, and Jared Colley, Chief Innovation Officer at Mount Vernon, to explore the skill shifts reshaping education in the age of agentic AI. Together, they unpack what happens when AI can execute, produce, and iterate — and human value moves upstream toward framing, judgment, and question design. From Vriti's Portrait of an Educator project to Jared's research on cognitive debt and question architecture, this conversation cuts through the hype to ask the questions that matter most: What kinds of minds must we form when execution is abundant and framing is scarce? Whether you're a school leader, educator, or lifelong learner, this episode will challenge your assumptions and inspire your next move.

Transcript

SPEAKER_00 0:02

Welcome to Start with Questions, the podcast where we explore bold ideas shaping the future of learning. I'm your host, Christy Lundstrom, CEO of Mount Vernon Ventures and head of school at Mount Vernon School in Atlanta and Mount Vernon School Online. Today I'm joined by two people whose work lives at the intersection of education, emerging technology, and what it means to be human in an increasingly automated world. And what excites me most is that both of our organizations are grounded in research. What are the data sets and the stories telling us? That's what we'll explore today. Jared Collie is our chief innovation officer and serves as editor and consultant at Mount Vernon Ventures. That's the RD unit for Mount Vernon. And Bredi Saraff is the co-founder and CEO of ED3, a nonprofit helping educators navigate emerging technologies with clarity and intention. The ambitious goal is to train and upskill at least one million educators on emerging technologies through research-based pedagogies. She's also the founder of K-20 Educators and serves in advisory roles with XPRIZE, Full Steam Forward, and Me Primer Bitcoin. Reedy, let's start with you. Tell us about your journey. How did you get to now and what are you working on right now?

SPEAKER_02 1:33

Sure. Lovely to be here. I uh have been in education for about 18 years. I started with Teach for America right out of college. And uh I joined Teach for America because I didn't know what else to do with my life. Um, and it ended up changing my life entirely. I drank all the Kool-Aid and I was a teacher for a bit, I was a dean for a bit, I was with the charter network that I was placed in for about 10 years in various roles. And at the very end, I was the network director for professional learning across 15 schools. Um and then I uh got this opportunity to travel around the world building schools, independent schools, um, where I was the global head of professional learning and we um teach, we taught educators um about different types of pedagogies, we coached school principals, I built the infrastructure for their evaluation systems, PD systems, summer institute, all of that. Um and I ended up spending a lot of time in Shenzhen, China, where one of our schools was, and in DC, and then also in India and all over Europe. Um, and then uh I started uh the uh well I started the nonprofit um in 2023, but uh it was uh 2020 that really brought out uh the sort of trigger point for it, which was when I was traveling around the world with this uh this this startup, I found that educators around the world were doing really incredible things, really innovative things, but they weren't sharing their best practices with each other. So there was a lot of reinvention of the wheel. And there was also very little technology showing up in the classrooms, even in the most well-resourced classrooms. And so in 2020, when everybody was trapped at home, I started um thinking about well, how do I connect educators in a way that uh leverages technology and allows them to share best practices? And so in 2021, I hosted um, I believe it was the world's first um avatar-based conference for educators. Um, and it was a choose your own adventure avatar-based conference where educators actually got to choose how they wanted to experience learning throughout a full Saturday. And I knew that it had been impactful when after eight hours of the conference and I wanted to shut it down, um, educators just refused to leave the platform 11 hours in. And at some point I was like, I need to go to bed. So um I had to shut it down. But um that basically helped me understand well, there's this hunger for you know, sharing of ideas and best practices across the world and learning about emerging technologies in a way that really combines pedagogy and thoughtful um practical practices. And so um that is what helped launch uh ed three. And um yeah, what we do today is uh we have three main areas of work. We have an RD arm where we try to push uh best practices and actually really innovative practices for the next frontier of education. And uh one of the projects we're working on there is called The Portrait of a Teacher in the Age of AI. We have a global community of practice that started from that first conference. We have about 4,500 educators around the world that share best practices with each other around emerging tech, and we give microgrants to educators to facilitate peer-to-peer learning events, and we've seen uh events all over the US, um, but we've also seen them all over the world, like in Nigeria and India and Jamaica and Columbia and Australia and other places. And then we have a PDARM where we upskill educators on AI combined with pedagogy through online courses. And um, Christy, you asked what I'm working on right now. Well, all of those things, um, but my attention is very much on uh the portrait of an educator in the age of AI. Um, I spent a lot of time thinking about and diving into the research that we're doing right now.

SPEAKER_00 5:28

Thank you. Thank you for sharing. Jared, why don't you give us just a quick picture of where our work intersects with Reedy's and also what is ventures focused on right now?

SPEAKER_01 5:39

Well, thank you, Christy, and thank you, Vreedy, for being here. Um it's fun to be a part of this conversation. And yeah, at Mount Vernon Ventures, uh, right now we're working on what will be probably our next big RD report in our own RD wing, where the focus in the research has seemed to coalesce or begin to center around what are the major skill shifts that are taking place in the context of agentic AI? So even moving beyond the impact of generative AI and really thinking about how is the presence of agenc AI and ambient AI in our workplaces and our learning places, how is that going to demand new skills? What foundational skills will remain sacred and how will those interact with each other in a way that's not overwhelming us as educators as we feel like we have to continue to build more and more capacity in our students as the world continues to change with so much speed and complexity. So that's what our latest RD report is going to be about that's coming out probably in August. We're doing a lot of consulting work as well with schools around AI. Um a lot of schools are coming to us about assessment design in the age of AI. We've had schools come to us about like, well, what would a responsible use policy that's both student and teacher facing, what would that look like? How could we put that together in a way that um uh is relevant and such a quickly moving uh uh uh area of emergent tech? So so we're doing a lot of work around AI that I think uh intersects with Reedy's work in in really complimentary ways.

SPEAKER_00 7:13

So let's jump right in. Reedy, there's so much noise about AI. We've just mentioned it 12 times in the last 30 seconds, and everyone's talking about it, and sometimes people don't even understand exactly what we're talking about. So there's hype. But what is actually changing and what isn't? And you have worked a lot about thinking where does the human decision-making power still matter most? So can we start there from where you see it? What is changing and what are people getting wrong?

SPEAKER_02 7:44

For sure. Um I think, and and this actually comes from research that we're done, we've done from the Portrait of a Teacher project. We know that adoption right now is nearly universal and it's happening much faster than training is happening. And so we're seeing a lot of use of AI from teachers and from students that isn't necessarily transforming practice or making things better. It's just using AI to the capacity that AI can be used right now. And what I mean by that is um, you know, if you look at the large language models, there's a lot that you can do with it, but right now it's mostly used as either a search engine or um something that helps you produce, you know, things for the classroom or um for assignments, things like that. And then ed tech tools are really all about meeting the needs and the demands of teachers right now, which is really about, again, the lesson planning, assessment, things like that. So it's not necessarily shifting practice, it's just allowing practice to happen faster and allowing you to do faster what you had already done before. Um, the the thing that we do see is that the administrative layer is being augmented and that is saving a lot of time for a lot of leaders and teachers, and that's a really good thing. Um, and then we know that um, you know, universities right now are rushing to credential AI literacy in different capacities. And so um, you know, that's resulting in a lot of focus on tool use rather than focus on what you just mentioned, which is discernment and some of the um, you know, core skills around cognitive and relational capacities. And so I would say um, you know, that there is, you know, a lot of adoption, a lot of sort of like progress. But I think the thing that we could be rethinking is that a lot of us are thinking about the rapid integration of AI into schooling and learning systems and things like that. But because it's moving so fast and because the technology is really an infrastructural technology that can really change the way that our systems operate, I think we're missing the bigger picture, um, which is that we really need to look at the pedagogical integrations and the idea that um it's not that we should be thinking about whether uh, you know, how to use or or you know, whether AI should be used or not, because AI is going to be a part of everything we do. It's really about like when to use it, um, when to decide when uh uh to to to use it, how to um discern the outputs that are coming out of it. Um, and and then fundamentally, this is like a much bigger question and a much harder question, but like what is the purpose of learning and schooling anymore if AI can do so much of it for students and for teachers, right? So like I think that's one of the biggest questions that I think we should be asking ourselves. And it's like a really incredible opportunity to ask ourselves because never before have we had such powerful technology that can do so much for us. Um, and then the other thing that I would say, aside from sort of like rethinking the purpose of schooling and learning, is that right now there's a lot of um use of AI that is really focused on the short term, so on short-term performance gains, on short-term, you know, learning outcomes. And, you know, there have been a lot of studies done all over the world where people say, Oh yeah, you know, I I gave an AI tutor to my students and this is like the the outcome. And I'm so happy that like there's so much learning happening. But like, is that assessment you know valid any longer? You know, is that going to be a good proxy for learning any longer? And then um similar to that, you know, when when we're thinking about AI utility um related to efficiency gains, are efficiency gains really a good use of AI, right? Is that an AI positive thing or an AI negative thing? You know, is it really going to improve the competencies that we really want um students to have? So I'd say there's um, you know, uh a big sort of, you know, uh mix of things that we could be thinking about. Um I would say the biggest question, the most important question is that we shouldn't be thinking about how to use AI. We should really be thinking about um what the purpose of schooling is.

SPEAKER_00 11:58

Yes, thank you. I I hear you talking about the portrait of the educator project. Could you give us just a little bit of quick uh discussion or definition around that? Who is it for? What is the scope? I don't think we're all familiar with it. For our listeners, can you orient them a little bit?

SPEAKER_02 12:15

Yeah, so it's a multi-year research project that um is investigating the question how does the role of a teacher change in a world uh transformed by AI? So we're not necessarily trying to identify what role does AI have in a teacher's uh, you know, uh persona. We're trying to identify how does AI actually fundamentally change, you know, the way that teachers are going to interact with students and what their role is gonna be. And so we have um a bunch of different research strands within it. Um one of them is called um between promise and practice, where we're trying to identify how is AI showing up in the classroom today, what impact is it having on teachers, and then what type of um ed tech ecosystem is being built for teachers. So, like what kind of features and tools that are AI related are are being built for teachers so that you know we know what kind of interaction they're gonna have with tools. Um, and so we're doing a bunch of surveys of teachers, of leaders to try to understand that. Um, and then the second uh research strand is called the architecture of the teacher role, where we're trying to identify what are all the things that teachers are responsible for today. And if we were to put AI into the mix at various levels, like, you know, let's say all AI versus some AI versus no AI, um, how would that role change? And then how could we configure different um responsibilities of teachers to create several different personas that could create a team teaching model that could really serve students well? Um, and that team teaching model would include AI team members. Um, and then the third research strand is called um the science of artificial relationships. And um this one just got funded by Gates. And the reason why we pursued this one is because um AI companions are becoming a very important um tool for young people. And um AI companions are one of the largest growing uh you know type of tool within the AI ecosystem. And so we know that AI companions are going to have an impact on cognition and on learning, but also on the relationships that teachers will have with students. And so we're doing, uh we're investigating how will AI companions potentially change the teacher role as well. And so we're bringing all of these research strands together to create open source frameworks and tools that any state, school district, school can use, graduate school can use to create their own portraits of teachers and personas of teachers. And then we're also creating a second set of tools that will allow them to audit their evaluation systems, their hiring practices, and their professional development practices against whatever portrait they created so that they can actually backwards map to their portrait of a teacher or their personas of teachers. And so that's sort of the ambition of the project that we're hoping to um complete the first phase of uh by uh January of next year.

SPEAKER_00 15:10

That sounds exciting. Thank you for sharing. Jared, you've been doing signal scanning and research around what you're calling the skill shifts. And so it's not that far from what Vredi is describing. And for our listeners who may not be tracking with the term agentic AI, I heard Vreedy mention that. She was also talking about how we might have AI partners on the team. She was talking about AI companions. But if we get a quick grounding and what does agentic mean in this context, and why does it matter for educators in the work adventures? And then I'll give the same question to Vreedy.

SPEAKER_01 15:44

Great. Yeah, so to put it simply, agentic AI are systems that don't just respond, but um they can act across multiple steps without additional prompting. Um, they can make decisions, they can execute workflows, and they can do so with quite a high level of complexity with a lot of steps involved. And so this is a really exciting moment, I think, because it's forcing our mental models about AI to shift. And what I mean by that is um I like to use the analogy of the of the video camera or the film camera, in that when we first started making films, we still made them like stage plays where we just set up a camera and we film people acting on stages. When AI first showed up with Chat GPT, I think a lot of us were using it as a search engine, as a retrieval um uh information base. And we were uh then starting to realize, oh, it's got some really cool interactive capacities to it, but agentic AI is really on a whole nother level in terms of how we must think about what AI's role is in our schools, what it is in our organizations. I'm reading a book right now, actually, just to build off of what Breedy said about efficiency gains, you know, how can we go beyond that as we think about how we interact with AI, how we deploy AI, how we use it to add value. There's a great book out called Reshuffle. And in the book, the author says, you know, if you're thinking about AI in terms of efficiency gains, if you're thinking about AI in terms of automation, if you're thinking about AI even in terms of just augmentation of individual performances, you're getting it wrong. He thinks that the author of the book thinks that the most powerful thing that we should be paying attention to about AI is coordination and AI's ability to coordinate knowledge work. And his analogy, his historical analogy is he tells the story about Singapore and how Singapore became so wealthy as a nation. And it's not because they owned a lot of resources, it's not because they were producing and manufacturing a lot of stuff. But what they did is they led the coordination efforts of standardizing shipping containers for global trade. And by standardizing shipping containers for global trade, we then had to standardize railroads, we had to standardize the size, say, of a tunnel so that the shipping containers consistently could pass through. And it totally uh uh rearranged workflows, if you will, of global trade. And so the real power with agentic AI is how can we tap into its power to coordinate knowledge work? And this is gonna lead to a lot of shifts in the workplace in terms of skills, much like Freedy is exploring with the portrait of an educator. It's also gonna lead to skill uh shifts in skills in terms of what we are fostering and cultivating and building capacity for in our students. And I can just give you a few examples of where we might see some of those shifts happening. One that I've identified is a cluster of skills that we might say are shifts from production and execution as an individual's uh uh uh sole task responsibility to the role of direction, to the role of executive functioning. We recently wrote an article at MV Ventures about more specifically one of those skill shifts, which is moving from task execution as our sole focus to task stewardship. And stewardship works both ways. It's stewardship in terms of having responsibility over the management and the supervision and the care of how we deploy our AI technology. But stewardship is also about stewarding human skills and making sure we're not cognitively offloading or de-skilling to a point where it's harmful to our development and growth. And so stewardship is a very complex uh uh concept that we really need to pay attention to. Two other shifts that I'll just mention quickly. I think we're moving from this idea of authorship and ownership of work and moving more towards a um uh process transparency, accountability as how we show um um the validity of our work, even if it's our work plus a machine and other people and other teams and a more collaborative effort. And then lastly, I think we're moving from in schools and in the workplace a demand on performance of knowledge to a demand for performance of judgment and discernment as we start to think about how we might uh steward these AI systems and steward the development of human skills as we're making these shifts.

SPEAKER_00 19:58

Breedy, when you hear Jared talk about those shifts, what resonates with you? And do you see any overlap in what ED3 is hearing from the educator that you're working with and the research you're doing?

SPEAKER_02 20:09

Yeah, there's a lot that resonates. I think the discernment piece for sure and the the um focus on process. Um, there are two things that I I do want to, you know say about those things, which is um you started off with uh the definition of agentic AI. And I think what people I think agentic AI has a lot of potential. And I, you know, my husband and I have been experimenting with OpenClaw and building our own um AI agents, and um, we found that the technology is still pretty nascent and it's actually really volatile, not because we haven't made a lot of progress on it, but I think it's because a uh AI in its fundamentals is just such a black box technology that it's sometimes hard to actually decipher what's happening in it. Um, even, you know, companies like OpenAI or Anthropic are are still having trouble figuring out like what is being done with the information, just because deep learning is um, you know, goes through so many different neural neural networks and neurons to actually be able to produce an output, it's hard to understand that. And so um I give a presentation on AI agents um across many different conferences, but the thing that I always land on is AI agents are have a lot of pro uh potential. And um, you know, they're they're really fascinating because you have to build like a soul into it and a heartbeat, and you have to build like a principles set and everything. It's super interesting, but it it's not going to be impacting, you know, the layman um this year or next year, I don't think. I think there's like a lot of progress to be made in order for us to be able to actually maneuver AI agents in a way that is um safe and productive. But yeah, so I I think it's uh I think AI agents, again, are have incredible potential and I'm experimenting with them myself, but I just don't think that they're going to enter our ecosystem with an education like too soon. And then the second thing I wanted to say, um, Jared, about um all the brilliant things that you said is that um I think there is still room for the thing the the things that you mentioned are sort of like the before versus the after. Um I think there is still a lot of room and validity around, you know, remembering facts and details or understanding what's happened in history or recalling things because I think especially at an earlier age, say like K through five, um, it's really important that we build those foundational skills before we introduce AI to students. And um we we want them to understand the process of learning, um, have the cognitive friction, to be able to um, you know, recall facts, to be able to build the um, you know, internal systems in order to be able to like do the further like algorithmic work. Like, I I I don't I I totally agree that things are changing in terms of what's important, but I do think that still has validity, and I don't think we should just let go of it.

SPEAKER_00 23:18

So I want to grab you guys both because I want to talk about some some uh words y'all are using. Could you define cognitive friction? And then Jared, I'd love for you to speak to cognitive, like de-skilling cognitive debt. So if y'all could define some of these terms for our listeners and then we'll continue.

SPEAKER_02 23:34

Sure.

SPEAKER_00 23:34

Go ahead, Rivi.

SPEAKER_02 23:35

Yeah, so cognitive friction is just productive struggle. So what are you um struggling with or what what are you being challenged by when you're actually going through some sort of learning process? Um, and uh when we talk about cognitive uh debt or cognitive offloading, um, what parts of those are being offloaded to, let's say, an AI system and what parts of it are you retaining? And and then that goes on to build the foundational skills that you actually have in order to build further skills down the road.

SPEAKER_00 24:06

Yeah, I think it's great for us all to be reminded of the neuroscience at play here. And Jared, you also talked about a concept in your research and you mentioned it earlier on de-skilling. Can you talk about what that might look like in the context of students? And then we'll ask Breedy maybe to think about that in the work our teachers are doing.

SPEAKER_01 24:24

Sure. Yeah, so de-skilling is, you know, anytime that we stop practicing deliberately, some skill that we've learned, and by way of the absence of that deliberate practice, we start to lose some of that, um, some of that deep knowledge and some of that muscle memory around that skill. For me, um, all I need to do is go to a calculus class right now, and then I will realize, oh my gosh, I don't know how to do some of these problems anymore. And because I haven't really engaged with higher level math, you know, on my professional world or in my personal world on a daily basis, I don't get the deliberate practice that I used to get. Because skill development takes deliberation, it takes habituation. And by habituation, it means we repeat and we repeat and we repeat. And so if you're not doing that deliberate, habituated practice, um, some of those skills can go away. But I would like to say, not all skill obsolescence is a bad thing. Um, there's a lot of skill obsolescence that moves culture and moves society and moves history forward. Um, not too many of us uh know how to basket weave anymore, and that's probably okay. But then there's probably some what we might call um, you know, essential skills or catalytic skills or capacities that are so sacred and so part of what it means to be human and what it means to thrive as a human that we need to be so careful that we're not providing opportunities of frictionless environments where kids don't get that deliberate habituation of practice around those capacities so they can build them up and flourish in a world that's complex, that's challenging, and will demand, as they're saying, due to AI and due to it making the world more frictionless, it's going to demand even more higher order thinking out of our young learners as they enter this world. So, all the more the research is saying, all the more reason that we do need to gauge these kids in building the foundational skills so that they can exercise competencies to get them to a higher order level of thinking as AI does begin to take some of the extraneous loads, some of the extraneous things away from our uh workflow.

SPEAKER_00 26:27

I'll pause you there because I heard Vreedy say when you were talking earlier, introducing yourself, you spend part of your time upskilling educators on technologies. And this is interesting. We're talking about de-skilling and upskilling. What resonates with you? What are you grappling with right now? And how do you help our teachers think about where to preserve productive friction and where to let go?

SPEAKER_02 26:50

Yeah, um, I I think uh what Jared just named is super valid. And um, you know, we think about it that way too, which is like, what are the things that are worth preserving versus what are the things that aren't? So I think there's this like very, very common trope right now of how like, you know, if essays are can be written by AI, then we shouldn't do essays anymore. And we disagree with that. Um, we think the um the ability to be able to, you know, generate ideas, brainstorm, be able to write, be able to express yourself without the support of AI is actually really important for um students and for any adult. Um, and so it's uh so things like that, we actually um look at the sort of like the cognitive learning behind it, the learning science behind it, and like what type of competencies are being built through that activity that we can't let go of. And then, you know, once we actually develop that foundational skill, then you know, in the future, if you are using AI to help you brainstorm and um, you know, generate ideas, that's okay. And even like refine your language, that's okay. In fact, that's encouraged because that's how you know professionals today are writing most of their content. They're you know, turning to AI first. But in order for us to um build that foundational skill, we have to actually allow students to have that struggle first. Um, and then the other thing that um I think is really important that we preserve, and this is what we talk to our teachers quite quite a bit about, is a relational and metacognitive layer of all of these things. And so, like, how are we thinking about thinking? How are we thinking about the learning? Um, we talk a lot about um key competencies that need that we want students and teachers to preserve um throughout our PD and throughout, you know, this era of AI. And um we split them up between relational competencies and um cognitive competencies. And so for the relational competencies, we have all of our SEL things that we always have, right? So ability to build meaningful relationships, ability to respond to others and yourself and your own emotions, ability to manage your own emotions. But then where AI comes in for the relational competencies is um the ability to respect and set relational boundaries with both humans and AI systems, and then the ability to distinguish human signals from simulated or algorithmically generated ones. And that's where like the AI companion stuff comes in. And then for the cognitive competencies, it's the stuff that we already know um, you know, about uh building um the ability to think in different ways, critical and creative thinking, um, the ability to problem solve and decide um, you know, uh how to go about a problem, things like that. But then the where a the AI piece comes in is for the cognitive competencies is the ability to frame meaningful problems and decide what cognitive uh work to retain or delegate to AI, because AI is basically, you know, another, you know, tool that we can use constantly to basically, you know, do some cognitive lifting. And then the ability to understand and um how knowledge is produced um and limited by both humans and by AI. Um and this comes into play with like, you know, social media and with um AI slop and with um the content that AI is generating on an LM, how to discern that. And then the last thing is like how to actually judge that output, right? How do you discern and judge that output? Um, so those are those are some of the things that we um talk about a lot in our um PDs and try to um help educators sort of like, you know, differentiate when and when not to use AI. And then one thing that might be useful, and you're welcome to throw a link in um the notes afterwards. Um when we were uh exploring how to help educators um think about AI impact, we um started experimenting with the SAMR framework. Um and the SAMR framework, um, as you know, if the audience isn't familiar, it's basically a framework that allows, it was created by um Dr. Ruben Pentadura 20 years ago, 30 years ago. And it's not research-based, but it's actually been um very popular among the ed tech and educational ecosystem. And um, you know, S stands for substitution, A for augmentation, M for modification, R for redefinition. And basically it identifies how or helps you self-evaluate how technology is being integrated into your classroom and what kind of um task transformation there is having that's happening. With AI, what we found is that you know people are saying that efficiency, which is augmentation, which is the very beginning of SAMR, is a good thing all the time. But we actually don't agree. Efficiency is sometimes not a good thing. For example, if you're going to have um, you know, an AI bot grade 200 of your students' essays without you ever getting to know your students' writing, that's not a good thing, right? And um on the all all the way on the other end, which is redefinition of SAMR, um, you know, if you're redefining uh your learning completely because you're using technology or AI, what if that redefinition is leading to fewer relationships and you know um a lower emotional load that students have to experience through their development years, right? So there, so what we did was we added basically a matrix and we added uh another vertical to SAMR and we said, well, it doesn't actually matter how much transformation there's happening that that's happening in the classroom based on AI. It actually matters what kind of AI output or what kind of AI impact that transformation is doing. So is it AI positive where you're building cognitive capacities, you're building relational capacities, you're um creating student agency, you're um uh promoting um physical movement in the classroom, you're building relationships, critical thinking, all that. Or is it AI negative where it's allowing cognitive offloading, it's allowing emotional offloading, it's creating standardization, it's creating compliance and isolation. So those are the things that I think become really important in the age of AI that we haven't quite thought about before.

SPEAKER_00 32:55

Yeah, that was a lot. Thank you. I you've got me thinking about so many different things, but I want to pull on one of the threads and Jared, one of the shifts you've been writing about lately is what Vredi just spoke to about doing the work to defining the work. And I heard her talking about who's gonna do what and the judgment around that. I've heard you talk about answer production to what you call question architecture. And if you could unpack that for us in practice, what does that shift look like? Who's doing the work and who is defining? Because I just heard Vredi say that's something we need to help our teachers build the capacity to understand.

SPEAKER_01 33:31

Yeah, Vredi, I liked earlier in our conversation you talked about how um one of the skills that's in demand is our ability to frame, identify and frame problems in a productive and creative way, right? And I um we've been looking at Dr. Pringle's work, who wrote the creativity choice, is actually a school read that we're all going to uh be exploring this summer. And in her work, she talks about you know, creativity. We quickly go to solution seeking and solution making, as like, oh, creative solution maker, or you know, when in what she says where the real creative work happens is problem identification and problem framing, right? And our ability to do that. And that is something that humans still have an ability to do on such an intuitive level, on such a creative level, um, that it's going to be very important for us. I um and so yeah, I think when AI is getting so sophisticated at execution, when AI is able to do things in such frictionless ways, where the friction does come up, where the bottleneck happens is it's it's further upstream, right? It is problem definition, it is scope setting, it is putting in the right constraints, right? And then from there, like we said earlier, being able to play that role as a steward through the process. Now, in the context of school, I think we still have to, of course, emphasize the foundational knowledge, because the only way that these kids can be good problem framers, the only way that these kids can be good problem identifiers, the only way that they can do things like think critically is if they have foundational knowledge to think about, right? If they have foundational knowledge where they can identify problems that exist that are worth solving. And so just to echo uh Reedy's point about like the foundational uh uh sets of knowledge where kids need to know things is not going away. In fact, it's becoming all the more important so that they can do these higher order skills that are also being demanded. Um, so yeah, I think that it's going to be very important for us to bring kids into the process of not just performing tasks or executing tasks, but helping design tasks that are worth doing. Not just giving students prompts and driving questions and essential questions, but asking them to identify problems and formulate their own questions. And I think that there's all the more of an urgency around not doing one over the other, right? But instead figuring out ways that we can integrate these higher order activities resting on the foundation of what we know to be sacred knowledge that they need to have.

SPEAKER_00 36:06

Reedy, when Jared is talking about that, and one of the things we try to do with this podcast is equip school leaders, equip PD leaders, equip chief learning officers, curious teachers. This is the part where we say, like, what question matters? We're a start with questions podcast. What is the question right now, in your opinion, that we should be asking that we're not? I heard you say it had to do with purpose earlier. Um, and maybe you want to hold on to that. But if there's another question you want to encourage school leaders to be asking right now, we'd love to hear it.

SPEAKER_02 36:40

Yeah, I mean, I will start off with that earlier question, then I can go into other ones. But um, what is school for is a really important thing. So um the sub-question there is are we using AI to perpetuate the status quo, or are we asking what education should be for now and what are we building? Um, so I think that's a really important thing. Um, but I think what we can also ask sort of in the near term is um, you know, what does our curriculum assume that is no longer true? So, what are some things that um AI uh can do that, you know makes the proxies that we've created for learning and education no longer valid? Right. I think that's a really important one that we can actually ask tomorrow, right? And and be able to evaluate. And then the other thing um from the student side is um how are students right now interacting with AI and what is their capacity to understand their own interactions and their own thinking. So that metacognition piece, because I think um, you know, this happened with social media. Um, we were all very much influenced by technology, and it had an impact on our mental health, on our well-being, on our self-perceptions, on our awareness of others in the world. And, you know, it took us a while to understand how youth were impacted by it, but it took even longer, and I'm not even sure we're there yet, where youth understood their own um, you know, impact. And so if we can start asking that question now about like, well, how much do youth actually understand how AI is impacting them today? And what can we do to actually bridge that gap and not necessarily solve any problems, but just get them more aware of how it's impacting them, then I think that might lead to some solutions or some um interesting uh more questions that we might be able to find solutions to.

SPEAKER_00 38:32

Yeah, I'll share a reference in the show links as well. I heard a presentation at the conference, Brida, you and I attended last week from um Project Tomorrow, the Speak Up Project, and they've interviewed 65,000 youth. And the research and the data that is coming out of that, I think is attempting to answer some of the questions you're asking. And I think one of the most powerful takeaways in that session were here's what the data says, go ask your kids. Go ask the kids in your context, go ask the kids who live in the world that you live in, or don't, right? Both are so interesting. So I appreciate you you bringing that. And I love the idea of what are we assuming right now? That's good. Jared, how would you answer the same question? What matters most right now?

SPEAKER_01 39:20

I was, you know, uh very similar question to what Ridi asked. And I guess what I what I want people to think more about as I think about schools I've been consulting with and the questions they're asking, and then where I'm maybe I'm trying to push them to ask a little bit of a more um provocative question. The one that I'm thinking a lot about is like what assumptions about what teaching and learning should look like are being called into question due to the shifting landscape. Um, that's not just AI. It's so much more than AI. Uh I think that we have to put it in terms of all sorts of complexities and uncertainties that are shaping our world right now. I think a lot about how it gets us to think about intelligence is not this individual thing locked up in our heads, but it's distributed across systems. And so I think another really important question to be asking is how are we uh uh uh getting our kids, getting our learners, and getting our faculty and getting everyone in our organization to think in terms of systems, right? I think systems thinking is so important. And I also think that as this technology grows more ubiquitous and as these kids once again are subjected to so many screens, I think it's really important that we look towards nature for other forms of intelligence and for other forms of collaboration that our students need to engage with and connect them back to the intelligence of our planet. Um, and I think it's really interesting that with the advent of neural nets and more people understanding what, you know, how this technology, the architecture of this technology works, we're also hearing a lot more about the intelligence of forests as networks. We're hearing all about the intelligence of slime molds and so many different things that we've got.

SPEAKER_00 40:58

Slime molds, kids will love that.

SPEAKER_01 41:00

And so I think that this is a great moment to remind kids that we live in an ecosystem. It's technological, it's natural, it's human, and it's more than human. And I think it's really important that we can all start to uh uh instill in ourselves and our students what I'm calling like ecosystem literacy. Um, and I and it involves technology to nature. So that's one way I would start to answer that.

SPEAKER_00 41:25

Thank you. So I always want these conversations to go from sort of inspiration to action. Today is Monday. What should they do tomorrow on Tuesday? What would you say to a school leader listening right now? You've given some examples. One thing, what could they do tomorrow?

SPEAKER_02 41:43

I think uh I'll I'll say two things. One is ask your students what they're using AI for um so that you can better understand uh the broader picture of where AI can go for your school. And then the second thing is um connected to the portrait of a teacher project, start thinking about what the purpose of a teacher's role is and how they can best support students, given that AI can be another team member. Thank you. Jared.

SPEAKER_00 42:14

Um two things at the most.

SPEAKER_01 42:16

Two things at the most. Oh gosh. All right. Start auditing your assessments. I would start auditing your assessments and think about are these tapping into what we know to be human capacities and the kind of human capacities that makes this work worthy and makes this work purposeful. And so one thing I would encourage people to do is just think about what are your assessment systems rewarding and incentivizing? And is it the right things? Um, and so that's one thing I would I would mention. Um, since I only get two and Breedy said some good ones, I'll just say you should connect with organizations like Ed3 and Envy Ventures. That's right. And let's think about these complex issues together.

SPEAKER_00 42:56

Yes, translating that research into practice. And before we go, I want to hear from both of you what's next. What are you reading, thinking about, or working on that has you energized? If you haven't already named it, where where should we look? And and maybe, Ridi, you want to tell us a little bit about how to connect with Ed 3 work.

SPEAKER_02 43:15

Sure. Um, so the Portrait of a Teacher project is taking up a lot of our time for good reason because we really want to make sure that educators are prepared for what's coming next and then also um starting to see the evolution in their own role. Um, and so if anybody is interested in that, you can go to our website ed3global.org. Um, and then if you click on the RD button, you'll see all the information there. Um, I'll be honest, um most of my time is spent uh figuring out how to feed my child and uh get her to sleep. She's 15 months old, and I have another one on the way. And so um a lot of the books that I'm reading are about parenting. And so um uh Jared, I actually do want to say I love what you said about um ecosystem. system learning um and systems thinking because that is um something that I think about a lot. I talk to my husband a lot because he actually got his degree um in uh you know philosophy and systems learning and things like that. And so when we think about um you know the way that the world is connected and you said um I think the the uh I I forget how you phrased it something about planets and and smart planets and things like that. Like those are the things that I think we really need to to think about. And one of the things that we're doing um with uh this organization called AI for equity is this project called HQIM.ai and what we're doing there is we're actually um identifying what are some standards and some learning um out uh objectives that we might want to integrate into different uh grade levels that might prepare students for for the age of AI and our contribution at three's contribution is um the K through six frameworks and what we're doing there is um we actually believe that up to grade three um students should not be exposed to AI or screens and things like that but they should learn about the fundamentals of AI and so the way that we're integrating that is we're creating these um you know uh these standards and these um integrations into curricula that identify um ways to teach students about neural networks about patterns about anthropomorphism and they're all through what Jared talked about which is you know um nature and about systems that um exist in nature and then basically connecting that to how AI operates.

SPEAKER_00 45:36

Oh I love that very very helpful I will include that link as well in the show notes. Jared um just tell me what's next for you what are you thinking about?

SPEAKER_01 45:44

Well doing a lot of research like I said on our next RD report. So I think a lot of what we talked about today is um this was fun for me because I'm getting to rehearse a lot of the things on processing right now. But I'll give you one thing that I read recently that I just think is worth a read. The Burning Glass Institute put out a paper I think back in February called Which Skills Matter Now a data driven framework for K-12 in the age of AI. If you haven't read that one I recommend it's good and it's a great framework much I I was thinking about it when you're talking about SAMR. They have this four quadrant framework where one quadrant is low automation, low augmentation so that's anchors these are anchor skills, anchor capacities they need to be done by humans. And then above that you've got like high augmentation but low automated automation and so these are like places where we need to deepen our relationship and our partnership and our collaboration with AI. High augmentation high automation is the transform quadrant where we just need to totally rethink workflows of how we do these things. And then the bottom right quadrant is um streamline which is high automation but low augmentation. And so these are things like you know grammar check that AI can probably do and we can offload without too much worry about damage to the developmental mind. So it's a great framework I recommend looking at that piece. I really enjoyed it.

SPEAKER_00 47:08

So thanks thanks for the tip we'll include that link as well and to our listeners come visit us. Come to Mount Vernon we'd love Vreedy for you to come we were trying to build a culture where curiosity leads design is a daily practice and joy is the measure and if you're asking big questions as Vreedy encouraged us to do about purpose and pathways in the future you'll feel right at home here. You'll be able to find all of the links mentioned today including Breedy's blog, Jared's latest article in our show notes. And finally grade schools don't start with answers they start with questions. Thanks everyone