Why Schools Don’t Need Another Silver Bullet | Jennifer Husbands

People in education are they don’t have the move fast and break things mentality. No. Which I think is good. We’re working with children.

Yeah. We don’t want to break children. All right.

If I had to choose between the the space moving too fast or too slow, I would choose too slow. We don’t really want to experiment on kids entirely. Welcome to another virtual edition of AIE Edu Studios.

I’m in the virtual studio in my bedroom, but we are live virtually with Jen Husbands. Jen, I’m going to give you a chance to actually share your a little bit about your story because you’ve done so much and you’re very very in demand right now amongst folks in the space to just like just want access to your smarts and and also just like all the different experiences that you’ve sort of like brought you to where you are today is like you know I’d say one of the sort of pre-minent you know advisers to philanthropy and to education leaders but this is very much like an interactive conversation. We’re going to be, you know, I think kind of brainstorming together because I think you and I both bring a very complimentary but different both perspectives and expertise that we need to find a way to sort of marry if we’re going to solve this big question about AI and future of education.

But before we even dive into all that, I actually just want to show you some stuff. So, we may we may cut this portion. But have you so so have have you played around with vibe coding much?

Have you heard of VI? I’ve heard of it very vaguely, but I have not engaged with it. And I’m curious like is it coming up in conversations that you’re having with folks like when you’re in rooms talking about the future of AI and education or people talking about vibe coding?

Is it you know I think Yousef brought it up with me at the EFSG meeting. I think I think it was not in the larger conversation with everybody. I think it was just the two of us and I think he mentioned it.

It was him or it was you. Somebody said they’re going to be talking about it at South by Southwest. Is it you?

Oh, it’s me. Okay. Okay.

But he mentioned it too in that conference room in that like weird auditorium room. He mentioned it. That was the first time I’d ever heard of it.

It was either him or Bob. Now I can’t even remember. But it came up in that room and then it came up again with you.

So vibe coding is I’m actually going to show you. Let me screen up. Take a feeling and turn it into code.

Take a feeling and turn it into code. Literally, that that is vibe coding. It’s what the name would suggest.

It it is exactly what the name would suggest. So before I actually show you vibe coding it’s worth a a quick primer. So yes essentially it’s using natural language you know English in our case and having the AI basically take what you’re describing in natural language and generate you know anything you can generate with code.

So that could be software, that could be you know websites and this was sort of like a it was it was quietly something that was happening almost in parallel with you know we had ChatGPT that came out in you know at the end of 2022 and then we had GBT 4 that came out in March. And so you know the education sector was kind of like scrambling over oh like what do we do with all these chat bots and you know how do we build things on top of that like leverage these like chat bots or integrate them you know into different tools and meanwhile you know folks in in the tech sector were really what they were focused on was you know this thing is really good at writing code and they were using it to you know make engineers more productive. Initially it was almost like more like autocomplete if you can imagine that.

So you’re like typing some code in it sort of like will suggest sort of like you know you just tab to enter and it sort of like helps you write the code like that. And and the capabilities of vibe coding have continued to move forward and then as and then basically what happened is you had GPT 3.5 to GPT4 and it was I mean do you do you remember did you use 3.5 like the 3.5 I have not been a user of GPT really at all. I use I use Gemini sometimes for work.

So the extent of my experience, yeah, I mean if you if you use GBT3.5, what you would have it was like very cool, but it was it was like it was like a high school student level, right? So it can like write poems and it could write essays and it was it was cool mostly because it’s just like the first time you could see just like AI almost like engaging with you as like a another sort of sort of independent agent. And and it did generate stuff.

And then we had GBT4 and that was like crazy powerful compared to GBT 3.5. You know it was like now suddenly you’re talking to a you know college professor or at least like a you know like a master student. And I think that’s where everybody starts to realize like whoa this is a really big deal.

And part of that was baked into this idea that well the the technologies if it in a matter of 3 months we went from GPT 3.5 to GPT4 significantly more powerful model you know who knows where we’re going to be in 6 months 1 year 2 years and so there was this like anxiety this like urgency around we need to just give people access to AI because it’s just going to get so powerful and we don’t want to be left behind and so there the paradigm I think has been really about you know how do we just like implement tools and give you get get kids hands on, you know, teachers hands-on. But then after GPT4, we had GPT 4.5 and GPT5 and then many other models, you know, like Gemini, it was like barred initially, then Gemini 2 and Gemini 2.5, and Gemini 3 and and for the for the non-technical user, they did get better, but they it started to get better at the margins, you know, like incrementally better. Certainly still had all the same problems that GPT 3.5 or GPT4 had in terms of you know it wasn’t reliable hallucinated and like made stuff up.

And there was a sense of like almost like stagnation in terms of like yeah maybe this is just sort of like roughly what we have. But in the background the coding capabilities of AI kept getting better. And like in and in particular last year Anthropic released something called Claude 3.7 Sonnet and this was an incredibly powerful tool for coding and it it really is what kind of like unlocked the almost like the Cambrian moment where where now half of all of the spend on AI right now is just for for for like coding purposes.

And and then in the summer of 2025, a bunch of companies including a Google, which this is what I’m showing you is Google’s AI studio. They updated the user interface to be a lot more userfriendly because before you could vibe code, but it was you were still sort of like in, you know, a coding terminal and you would see lines of code and it was just it was overwhelming to somebody who had never done anything with code before. So they changed the user interface and now you get it’s just very it’s very very user friendly.

So I’m actually going to walk you through I’m going to walk you through the the process of designing something and we’re going to do a little design thinking together and and I just want you to sort of think about you know something that just like a problem that you have dayto-day. This doesn’t have to be even workrelated. Like just to give you an example, someone was telling me they just they just separated and they have two kids that they take care of every other week, which makes it really hard to figure out what food to put in the refrigerator because you’re like buying all this food for your kids and then like it kind of goes bad.

And so he always he always like has to figure out like what what could I make with what I have in the fridge. And so one thing that we built together was like a like a recipe generator that would be like sort of help him sort of like plot out, you know, like what to buy and like, you know, how to sort of like, you know, organize it. But yeah, any anything that we could just like just right out of the gate kind of like try to try to hack on together?

I mean, I hate packing. Okay. I hate packing and I have to pack often as you know because I travel both for work and for pleasure.

So, like there’s got to be a way to make it less painful. I don’t know. Okay.

It’s kind of a boring boring challenge, but it’s a real one for me. So, you people can send me up lists, but I just Well, where the trip, where am I going? Is it a business trip?

Is it a personal trip? You know, what’s the climate like? What’s the weather forecast?

I’m always on like, you know, snap doing like a screenshot of Weather Channel to say, "All right, 4 days in DC, it looks like this." or, you know, three days in LA or five days in Mexico, you know. Yeah. So I’m actually going to use just like a just a regular LLM.

This is Gemini. To help us come up with the idea first because that’s always like, you know, I think it helps actually do some ideation. What what are some of like the what makes packing hard?

Like why do you hate packing so much? I think it’s I think it’s kind of an executive function problem because I think it sort of like requires me to think ahead and plan ahead for multiple days. And that feels exhausting.

So, I procrastinated till the last minute and then I’m stressed. All right. And let’s and so one of the things you’re seeing me do here is you see this drop down.

I don’t know if you’ve used you said you use Gemini. Yeah. So a lot of people have like you know when they’re like yeah you know I’ve used Gemini and it’s just it’s okay.

I don’t I don’t think it’s that great. And I said like it’s like pull up your computer. Let me see show me some of your your prompts.

And I looked at all their prompts and all their prompts had just this this drop down. They were doing fast. Mhm.

So fast is the the weakest model, right? Gemini also the fastest and the most cost effective. That’s the title.

But if you want to really see the the magic, you want to click on pro. Yeah, I only use it in pro. Oh, so you’re like, well, so you’re already, you know, because it’s tied to my work.

It’s tied to my suite. It’s tied to my suite. So I just use it.

I’ve only used it for work and I use it for personal. So I’m always in my my admin suite. So I have more tools there.

Okay. So let’s see. A visual outris outfit Tetris app.

I love it. A backwards plan or timeline generator. Yeah, this one I feel like when do you have to do your laundry?

Like when do you have to, you know, the vibe check packing auditor. Vision. I’m going to avoid stuff with like vision cuz like that can be a little more complicated.

Okay. I don’t know about this one. Sensory friendly event matcher.

I’m not too worried about that. Okay. Okay.

So, let’s let’s go with this backwards planner. The reason I’m not doing this the Tetris app is what it’ll just take it’ll be long. It’ll be it’d be cool, but I couldn’t do it in like 10 minutes.

Yeah. Yeah. Yeah.

Okay. Timeline generator seems like a good idea. So, I like the So, I like the backwards planner timeline generator.

Can you generate a starter prompt for me to drop into AI Studio to get to an initial MVP? Okay, I know what MVP Importantly, I don’t want to use any API connections for this initial build. I don’t know what that means, but I’ll take your word for it.

And by the way, we’re talking about AI skills, and I kind of roll my eyes sometimes when people are like, "Oh, we’re going to do like a prompt engineering workshop." I roll my eyes because I feel like the only thing you really need to teach somebody when it comes to prompt engineering is that you can just ask the AI for help with whatever it is. So it’s like I want to do I want to do write a prompt to do XYZ. You just say how do I help me write the prompt to to get to that end goal.

So it’s almost like if you if you are able to articulate the problem you want to help solve. That’s almost all you need to know, right? But there’s a bit more to it, right?

This like this intuition of like some of these you know some of the reasons I didn’t go down some of these other paths is because I’ve I’ve like experimented enough that I’ve built that intuition. Okay. So, we have our prompt.

Where do you put that? And then we’re going to go to AI Studio. So, this is AI.

Okay. Okay. Okay.

If you go if you if you log in initially, you’ll come to this place. It’ll be home. You want to go to build.

Okay. Which is actually where you’re going to describe your idea. And even this is a bit too much.

Delete that. Just call it a web app. So, it’s funny.

We didn’t say anything about neurode divergence or ADHD, right? And it’s like time blindness, executive function, and I’m like, well, yeah, that’s that’s part of it. Okay.

Then I click build. So, with Vive Coding, you’ll usually have pretty much all the tools look roughly similar. By the way, this is another Vive Coding app.

Okay. If you want to see an example of this, and looks roughly the same, right? You have your your chat bot, your your chat box here.

And then a preview. So, it literally will like code and build the tool or the application or the software or the website. And you’ll sort of see it in real time as it’s developed.

And so unlike if you’ve used like a a website template, have you ever used like Squarespace or Wix or anything like that? So in Squarespace, you’ll go and like actually click and you can like change the text and like it like a text editor. The way VI coding works, it’s almost like you have your engineer sitting right next to you.

And so you wouldn’t go in and change the title. You might say, I don’t want it to say vibe coding sandbox. I want it to say, you know, vibe coding playground.

And you literally just type that in there and then it would change it. Okay. Okay.

So we have, so this is where the preview is going to going to show up. It’s like literally doing the coding right now. Okay.

Creating some metadata. Okay, here we go. Do you have a trip coming up that we could I am going to South by Southwest edu, which hopefully I’ll see you.

So I’m leaving on March 8th, returning on March 12th. Okay. And you’re there.

This is a work trip. It’s called a business trip. Yeah, this is cool.

All right. So, we have like some like functionality. So, you can like check the box.

Okay. So, so this is now like truly the first the first output. And what happens now is you start to like iterate on it.

You say, "Oh, I’d like to be able to upload a my flight information and have it like sort of like add that to the plan." you might want to you might say, "Oh, during the like pack like for the like whenever there’s detail there for sure. Yeah. I want to be able to add like a to-do list underneath each of these sort of like like the sort of like the the cards on the timeline.

Yeah. And it’ll just build that stuff for you. Okay.

So if we took more time, we could we could pretty much build the app for you. Yeah. And I can send you this afterward if you want to hack.

Yes. Yes. I’m curious.

Yes. I want to play with it. But the Where’s my meter study?

It is here. So this is this is kind of what happened as I was kind of talking to you about. So, what you’re looking at here is, a a a a benchmark that, this organization uses to, figure out like how many times can LLM complete, coding tasks u, and how long does it take to complete the task?

Yeah. Mhm. And basically what we’re seeing is that the like the time horizon is increasing.

It’s getting longer. I would think it would get faster. I wish nation.

Ah okay. So, the the one of the big limitations with with with using AI to code is that are you familiar with context windows? Mm- So, let’s say you were going to write an essay or a an op-ed about, you know, something about education.

You could upload like background information to help it write. Over time the am the length of how much background information you could upload has increased. So it used to be that you could only upload like 400 words and then that increased to like 800 and then then you can upload like you know a six-page document and and now you’re able to upload like you know a whole book.

Mhm. Well, with coding, the way that that has played out is, you know, you could use AI to to to do tasks that would only take you, let’s say, 10 minutes or 15 minutes, but the minute you get to like longer tasks, like something that would take you, let’s say, 2 hours, there’s so much code that you’re writing in that 2-hour process, it’s all it’s creating all this context, and the AI just fails. So the higher on the chart it means like the the the longer the task the AI can successfully ah okay okay okay thank you for clarifying that’s impressive then and maybe a little scary for some people very very very much so so we’re we’re getting to you know pretty sophisticated tasks you know and and kind of my example of like you know Claude coowwork being written in a It’s like almost entirely generated by AI.

It has gone from a helpful tool to support and augment engineers to a tool that can that can almost wholesale replace Yeah. Work that an engineer would do. And again, this this happened middle of last year.

Yeah. Wow. So, a quantum leap as they say.

It is. And and the combination of that with the user friendliness Yeah. Of being able to literally just like type in something you want and you have this and you can now share it just like you would share like a Google doc.

You can publish this by the way as like an application with its own URL and you could share with friends or with your with your child or anybody that might want to use it. So I that’s mostly I mostly just want to share because that’s what’s on my mind is we we don’t like we we have to come to terms with this technology is moving way faster than the infrastructure of the educa of education and that I think that’s a segue to sort of like the conversation I’d love to have with you Jen is you know how how do we make sure that the education sector is actually thinking the long game and not just so focused on the the shiny objects, the tools of today as we’re thinking about like what does school transformation look like? What does the you know building capacity look like?

Is it capacity to use you know tools like magic school or is it something a little bit bigger that is reactive to whatever the tools might look like in one year and 2 years and there’s a whole separate thread that we may or may not have time for which is what are the implications for future work right if like these if like you know you can complete six-hour tasks now with AI what are people going to be doing like will they be doing anything is a question people are asking but before we get into all of that Jen could you give us sort of like maybe like a short ja through through your journey and like sort of what brought you to this place where you are today where you you you’ve been sort of pulled into these like really interesting you know highle conversations about sort of charting the future of education and go back as far as you want we can you know thank you this is fun I’m happy to see you happy to be doing this and there’s certainly a very long version of the story that I could tell that I’ll try to tell a bit a bridge version but I do think it’s important the roots are important right Like I grew up in the suburbs of Philly and my parents split when I was nine and my mom moved into Philadelphia. My dad lived in the suburbs, which is ironic cuz my mom was white living in a predominately black neighborhood and my dad was black living in a predominately white neighborhood. But I saw education and equity firsthand because of the experience of existing in these two neighborhoods.

The opportunities available to young people in those two settings were night and day, completely different. So I think that informed me, you know, through undergrad. That I was at first I wanted to be a journalist but I decid I ended up switching to focus on education based on some classes that I took late in college and then work experiences post I worked briefly in the private sector and then I worked for an Americanore program very eye opening opportunity I was able to visit all the sites in Philly and New York and be in schools and I was like okay this is it you know read sabotage equalities I just okay this is what I want to do with my life is think about this this work on this issue so I went to grad school and earned my PhD in education policy and administration at Stanford ended up doing my dissertation on San Diego a high school reform initiative that was happening then and I ended up studying that.

So I moved to San Diego and worked on that for a couple years postoc and then I ended up getting hired after doing a consulting project for high-tech high ended up getting hired there and so that was definitely a cutting edge kind of place to end up especially in 2004 I think it had opened in 2000 so you know explosive growth numbers of schools and also yeah just for our listeners who haven’t heard of it yeah can you give oh gosh itai to me itai I kind of call it like it’s the education’s perfect storm in some ways in that the the thinking behind it the founders, brilliant people who came together. They had an architect on the design team. They had philanthropy who was excited to support this effort in a big way.

They had land that was going from that was sort of being decommissioned whatever like a naval pl base. So they got like built you know a dollar one of those like you know they got land and space for a great price in a great location. And so it was like all these things came together and the vision was to create project-based learning focused schools public schools.

And so they started with one and then they added a middle school and then they added another high school and then and now I think across the sort of San Diego region I think there’s 14 schools or something like that. So continues to be an innovative place. When I was there they got approved to certify their own teachers which was the first charter approved to do that.

So that was part of my job and they had a large Gates Foundation grant to try to replicate back when we thought you could replicate a perfect storm. We have gotten a little smarter since then. But there were schools that were started in that high-tech high network that are still going strong and many have closed from that initial network.

You know, it would actually be interesting to study the sort of trajectory of that whole network of schools that we helped to start. And then the founder of High-Tech High, Larry Rosentock, wanted to start a grad school. He just was like, let’s I want that.edu.

Let’s start grad school. And so we kind of I had to figure out how do you start a graduate school in a K-12 charter context in California? And we did.

And that that grad school is still going very strong. But San Diego was not sort of for me ultimately. So I moved to Chicago and worked at Chicago public schools and then kind of moved around as many of us did sort of after Arie Duncan left Chicago and went to DC to join the first Obama administration.

You know things kind of got unsettled in the Chicago context but ended up leading a nonprofit and then going to the Gates Foundation where I was for six years on the networks for school improvement strategy. So I ran the community of practice for this national community of practice. There was about I don’t know 21 grantee organizations in there and so a real focus on secondary schools using continuous improvement moving particular outcomes that were sort of you know codified in the in the strategy working with teams that were working on these things doing capacity building convening stuff like that and so then my strategy was ending so my role was ending so the question was what’s next?

This is 2024 and I decided to start my own consulting practice you know kind of I’ve been doing adult learning work in education you know for 20 plus years so I thought you know and I’ve now seen it from philanthropy I’ve seen it from a district perspective I’ve seen it from a nonprofit charter been an ed of a nonprofit I’ve been employee of a nonprofit so I was like I think I have lots of a good almost 360 perspective on some of these things you know I think I can be helpful so I started my consulting practice which is called steward education adviserss in September of 2024 and then I met you at a convening that I was helping to plan for the education funer strategy group in partnership with ed council and the rest is history I’m here and here you are. Yes. Yeah.

And and yeah education EFSG education fun strategy group I mean this is I mean these are some of the most sort of sort of pivotal philanthropies in the education space coming together. I mean, I remember you and I met this was months and months before at Grant Makers for Education actually when you had mentioned and so there was a lot of planning that went into this and it it was reflected by the way hold on I have a cat who is wanting to get Yes. I mean I’m curious like where where are folks minds right now?

Like is is it is it excitement? Is it like are folks just sort of like still feeling a sense of like you know confusion or anxiety or do do people is there a plan? And I mean I just you know it’s been the three years on since you know chat GBT and language models kind of were thrust upon us like truly just from your vantage point where about AI about course of course tell you what people are feeling from my perspective in general in the world right now but with AI specifically you know oh boy that’s a good question I think that there is definitely anxiety I think with any any technology anything that moves really fast makes people anxious just cuz they’re like I don’t know what this is.

I don’t have control over it. There’s a lot of unknown there. I think that when you’re working with smart people, you know, then these thinking about these people in philanthropy, they want to be ahead of the curve and there’s no way to be ahead of this curve because the curve keeps shifting.

And so I think that sort of there’s like there’s almost like a humility that you have to be like, okay, I’m not going to be the person who knows the most about this, but can I know enough to at least be helpful? And so I think that there is and I think probably like there’s almost like an adoption curve of people right like there’s some folks who are like all right let’s do this let’s get on it there’s people who are like let’s wait and see you know there’s lagards there’s so I think there is definitely I I don’t think there’s certainly not one reaction I do think that there’s a pretty universal awareness that is a thing we need to pay attention to but I think there’s a lot of trying to learn and figure it out and see what’s happening I don’t feel like folks my experience is I don’t feel like folks feel like they have a good handle on it. And so, and it’s like a second step cuz it’s like especially in education if you’re thinking about, you know, what are folks doing, what do they need help, first you have to kind of understand the lay of the land if you’re going to think about where you might step in and support it.

And so, I think that things are still sort of they haven’t quite gelled. It’s a tricky time. I think it’s a tricky time and because that is laid on top of a world that feels like it’s in crisis and is actually in many ways, right?

And in ways that have huge impacts on education both structurally and students lives and communities specifically you know so like I think if this were coming at a time that there wasn’t all this other stuff maybe people would be able to attend to it in a more disciplined way but it’s like one thing amongst lots of things that feel very urgent right now so I think that is part of what’s sort of stinying having a more organized response to it yeah I mean I mean education is always hard because there’s like it there’s always crises. I mean, even if it’s just like, you know, school bus strike or, you know, a like, you know, my mom taught in Akan public schools and just and then she she was like an assistant principal for a while and like dealt with like discipline and just like, you know, the stuff that she had to deal with on a day-to-day basis. Setting aside like the macro challenges like even the day-to-day in good times is very complicated and it’s hard to be focused and then you’re layering in literal like you know it’s like budget cuts and then I actually just presented at a school on last Monday and there was like the the national walk out on the previous Friday in protest of what ICE was doing in Minnesota and you know it’s like I was actually there to present it was like the the educ the teachers staff were all together.

And so it was like a staff, you know, learning opportunity. And like the first 20 minutes were, you know, folks kind of like just debriefing on all the stuff that happened around that and like the way that the you know, just like how the teachers were feeling and you know like what it meant for for those teachers who didn’t want and it was just like just as a way to demonstrate you know we’re here to learn together. But that was a conversation that had to happen, right?

You can’t just say well this AI thing is really important so we’re going to just ignore the other stuff and funders I’m sure getting a lot of emergency appeals right now whether it’s organizations that are you know having to mobilize you know for what all the various you know needs that are that the moment is requiring you know from activists from nonprofit groups but then also just like because of budget cuts. Yes. Exactly.

You know, it’s like there are most funders that I’ve talked to are not in a growth mindset right now. Like they’re not thinking about how do we expand. They’re thinking about how do we keep folks lights on and you know and they’re getting hit by new tax laws that are changing their ability to give.

Yeah. It’s very hard. Yeah.

And yet we managed to get I think it was like what 40 or 50 funders to take two and a half days together and create that shared space. And it’s interesting you you I one of the things you said struck me. It’s like People in education are not like it.

They don’t have the move fast and break things mentality. No. Which I think is good. I I I think I would rather that we’re working with children.

Yeah. We don’t want to break children. Right.

So, yeah. I don’t know. I mean, I don’t know about you.

I think if I had to choose between the the space moving too fast or too slow, I would choose too slow. Even though there’s there’s risks with that as well, because yeah, we’re dealing with kids. We don’t really want to experiment on kids experiment on kids entirely, right?

And then you still have the adoption curve there cuz obviously we have schools, districts, places that are really seizing this moment, you know, like we had folks from Gwinnet County like they’re doing really interesting forward thinking stuff about how to integrate this in. We got micro schools, you know, doing there’s lots of, you know, so there’s there’s again the curve and then there’s a but I think that yes, the bulk of the curve is still on the slower adoption end of things for sure. I mean, this feels like a crazy moment and it probably is maybe the fastest technology has ever changed in the history of humankind.

I mean, I’m trying to think of maybe was there ever like a faster takeoff and I can’t quite think of anything. Maybe like the nuclear bomb in terms of how quickly that changed geopolitics, but you know, all the other industrial revolutions played out over like, you know, two to five decades. Not how would you compare it to the internet, the advent of the internet?

Well, I mean, think about I mean, I remember when my parents had dialup. Yeah. And I remember when we had broadband and that was like 5 years later.

Yeah. Yeah. And then I remember when I got my first smartphone slower.

The progression was a lot lot slower. It wasn’t 6 months from one, you know, release to the next. That changes everything or whatever.

Yeah. It was it was much slower. That makes sense to me.

Yes. That’s right. But still I mean like I think that’s like part of you know if you’re I’m a student of history and as you look at sort of like as we’ve had more like you know we’ve had a couple industrial revolutions you know under our belt this really does feel like an industrial revolution the time scales do get compressed each time because you’re in in each case you’re building on top of you know everything that came before.

So, you know, AI already had the internet and so Chad GBT was able to hit, I think, like 100 million users in like hour or something. Because everybody’s already connected and sharing it on Tik Tok and reals or whatever. Yeah.

But I the reason I bring that up is because this is not the first time that education is had to make sense of new technologies and figure out how do we sort of transform? And I’m curious of like did it feel like this? It’s like was there the same kind of like scramble and mix of excitement and anxiety about getting computers into schools or getting internet into schools?

Yeah, I don’t I think it’s I think it feels pretty different. You know, I remember the first Apple, you know, computers coming in when I was in elementary school. That’s, you know, where I am.

Gen X. I think that why this feels different to me, especially thinking about like edtech, like edtech was sort of about how to deliver content better, you I mean, it was sort of not fundamentally changing the the nature of things, you know, maybe some of the the how, but not as much the what. I think that this this is changing so many things.

It can change how you teach your kids. It can change how your kids respond and and produce content. It can change the kinds of jobs that are available to kids.

It can change what kinds of information and resources are available to kids outside of school or in school. Like, it’s changing. So many facets of life at the same time.

Whereas I think those other pieces were much more like circumscribed. I was like, "This is going to change how you teach math." You know, that’s still a thing and that’s still, you know, not not inconsequential. But this thing has so many dimensions to it and so many ways it can change our lives that I feel like it’s just it feels so much bigger.

Yeah. You mentioned Quinn County. You like there’s there’s actually like a really long list of schools that are, you know, forward thinkers that are actually doing like walking the walk and they may not have all the answers, but they’re like, "Hey, we have to try something." In part because the kids are sort of forcing them to, right?

Because like at the end of the day, the kids are using this using this stuff, you know, with or without you. And even if you block it and ban it and have all the academic integrity policies in in school, they can just go home. And so that’s become a trigger to schools are like, you know, they’ve been pushed over the ledge and and there are some like amazing stories and I I it’s but it’s not enough, right?

Like it’s it’s it’s there’s a long list, but relative to the, you know, tens of thousands of school districts across the country, it’s still a really really small segment of them. Yeah. What do you see as like the most important conditions that need to be in place for a school to to go from being sort of on the back foot to the front foot?

I think that I think leadership is huge. I think how who the leaders are, what the leaders sort of know their sort of like growth mindset and openness and then the way they sort of think about introducing you know new tools or or new approaches to their staff. I feel like that is sort of the crux of a lot of change.

You will have some bottom up places where you’ll have teachers, teacher leaders who embrace things and bring them to their colleagues or bring them to leadership. So it’s not always a top- down thing. But I think the role of the leader, if you actually want something to to diffuse across more than one classroom, I think leadership is really important.

And when I think about even when I think about my career which a lot of it has been about diffusing innovation you know it may not seem obvious you look at my resume but I mean working at high techch high was about how do we get more schools to use project-based learning you know even when I was at academy for business school leadership how do we get more schools to be thoughtful about you know really turning around the lowest performing schools you know some of what we did was probably I would feel differently about it now in the rear view but that was an urgent sort of push and coming from the feds and So I feel like what I’ve observed in an effort to try to disseminate innovation is that resources are only useful to people when they’re useful to them. It’s very simple sounding, right? So you can have all the resources in the world, but if I don’t see it’s got to be user centered at some level, right?

If I don’t see how this thing is going to help me do my job better, help my kids learn better, help me prepare my kids for the future better, then I’m going to probably not use it because we still use a closed door classroom. And so I think that I think and I even see this just in the general world, you know, like a lot of AI stuff is being pitched or like being put on my phone and I’m like I don’t want this. How do I turn this off?

Like I I don’t want it’s not helpful to me. It’s not because I don’t like AI or don’t want it. It’s just I don’t want that right now.

I don’t that doesn’t help me use this app better. So like if we keep foisting it on people then I think you’re going to not see the kind of uptake you want. I think you have to work with people to figure out what is the problem you’re trying to solve or how can we use these tools to help you do your job better or to free up your time to do the things that that are only being able to be done by a human, right?

Like I remember I can’t remember it was someone I think at the ESSG meeting that shared some data. I think it might have been Yousef about like teachers using a tool to do the first pass review of these essays and give the kids feedback and then the kids turned in the revision and then the teacher read the final versions or whatever, you know, and how the student’s grades went up. Now, I don’t know all the details of how that happened, but I’m just like, how can we think about the things that that’s so to me that is meeting a teacher’s need, right?

Like you can do the first pass on essay and kind of get rid of like the easy like grammat grammatical mistake like stuff like that. Do that. I don’t need to do that.

Right? You can do a computer do that. Get the kid to improve it.

Then let me get in there and give that human level of feedback that that probably a computer can’t provide. A computer doesn’t know you as a kid. They don’t know your whatever.

So I think that how do we make those kinds of cases to people that like it can make maybe more efficient may let you use your time in way just separating out the sort of uniquely human things that humans need to do and then things that maybe could be done using another modality and present that to teachers in a way that feels helpful that doesn’t feel you’re trying to take my job away you’re trying to deskkill me you’re try you know like I just I think there’s you have to kind of sell it and you have to sell it in a that feels user centered and I feel like kind of a lot of what I’ve seen doesn’t feel that way to me and I think that you know and I do think there is a diffusion of innovation even within schools like you can have you have your early adopters so like okay go to where it’s warm have those people pilot some stuff and then the kids start talking about this cool thing they’re doing in so- and so’s classroom and that teacher’s like wait what are they doing I want to do that too I want kids talking about my class like you know like it so I think it’s sort of like you know some people say a wedge and spread, you know, like you got to get in there and then you got to spread it. But just doing it completely top down or not making the direct connection to how is this going to make my life better, my kids’ life better, the world better, whatever it is, you know, if without that rationale behind it, I think it’s it’s going to continue to be a hard cell. We know that, you know, people don’t have time, they don’t have the resources, you know, like we know that we’re going into a challenged environment, particularly in public schools that serve students with the highest needs.

So, how do you take this environment that already is stretched very thin and allow people sort of incentivize people to want to try something different when they feel like they have no bandwidth, not even enough bandwidth to do what they need to do dayto-day. I mean, that’s probably not, you know, the most innovative thinking, but that’s just that’s where I think we are is that we we’re not selling it well. Well, yeah.

And I think you know the the startup founders would say well we are we are making the case and they’re doing all this marketing and no doubt there’s probably some AI involved in sending out all the emails and because the teachers I talked to it’s it’s not that people aren’t trying to show them how AI could help them. It’s like they’re being bombarded with hundreds maybe thousands of you know companies trying to convince them that this is going to be helpful. But I think what you’re saying is really important.

It’s it’s the like who is communicating that is very like is pivotal, right? Like if this is coming from another teacher, the level of trust there is it’s it’s it’s not even like comparable, right? Compared to like cuz teachers get pitched stuff all the time, whether it’s AI or otherwise.

Yeah. 100%. And teachers also get told to use a bunch of stuff.

That is actually like the practice of teaching is essentially the practice of sort of like operating within an extremely rigid and frustrating bureaucracy which has gone through you know with no child left behind. This sort of obsession with measurement you know at the expense of well project based learning among other things. And so it’s not so simple as well, you just need an administrator who is creating an AI initiative and making sure that teachers are all using AI because a savvy administrator and certainly the savvy ones we talk to will say well I could do this but I know that’s actually not the best path to success.

This needs to be at least partially homegrown and like like that wedge and spread is interesting like that teachertoteer. Yeah. Or it could be a coach you know I mean instructional Yeah.

Yeah. We see improvement, literacy improvement. You know, if you, you know, unpack what’s happening in like Mississippi, there’s a lot of coaching.

So, what would could be the role of kind of an innovation coach or an AI, someone who sits with you almost like a therapist, what’s going on? What do you need help with? What’s not working?

You know, and sort of then says, "Here’s some things that you might try. I can help you. Let’s do this." And then I think then you could see some uptake, right?

If there again it’s a human you know this I think for this whole thing to work especially in education there has to be a human interface you know between these major points like between teacher and student maybe between you know product and teacher there’s a human interf you know like between principal and teacher you know I just think that there has to be I think there’s a humanization aspect to it that is being perhaps missed and that all sounds sounds well and good but the the the challenge in education is just the scale right so you have what 50 million kids, 4 million teachers, I think 40,000 districts. I I used to have that this number down many, many thousands of districts and it’s all decentralized. You can’t just go top down.

Yes. For better and worse, frankly. And so it is it’s easy to go to to look at one school and say, "Okay, like we could sort of map out how this could play." But then if you’re trying to reach, you know, the the really large, you know, audience of communities that are going to need this support, it feels a bit intractable.

And one of the pieces of your work at Gates which is really focused on community of practice models and one of our like sort of strategies and theories of change which is you know going rather than doing breath like actually going really deep with like subsets of teachers. We have like our trailblazer fellowship. What we found is that it’s not just about that one teacher changing her practices.

It’s that that teacher is now connecting into a community of other teachers and then they are bringing other peers along and they are supporting their administration and like making the case because now they’re an exemplar that the principal wants to celebrate and it becomes less about how do we get AI into our school and how do we support more teachers to do what you know Stephanie is doing and you know our trailblazer cohorts are about 25 and they run for 10 weeks so there’s no way that we’re going to trailblaze our way to every teacher our big bet is that you’re not just teaching that teacher, you’re building that alumni network and you’re building these sort of models and case studies. But I I’m curious like what are the other sort of secrets to success of a community of practice model and this could be your advice to us or just sort of more broadly for like what the space needs. Well, I mean there’s I think there’s two pieces.

A community of practice is is a space is where you bring the people together, right, to learn, to share, all of that. But it’s it’s you have to have a network, right? I mean the network piece is really important.

So in this case then I would say I love the idea of having like teacher leaders within schools who can try to model and share and sort of demystify these technologies within their building. But then you need to network the schools together and have the teachers from across the schools talk to each other and then maybe you even have something above that. Right?

So I think your question about sort of how do we the scale question I think in my dissertation research even though we weren’t talking about AI but we were talking about improvement school districts school districts are a very important unit of change and I think you know like I know Gwynette is an example that’s often cited and I hope that there are others and I’ll be happy to cite the others once I learn about them but I think that if you have ownership at that district level and then you have activities like this happening at the schools and then you’re bringing them together to amplify it that is a big unit of change that then becomes an exemplar that you and I have both heard of that other people are going to visit and learn from and and get their documents and see how they’re you know thinking about it and they’re attracting funding and you know like so I think the district is a an important unit of change of course some districts are one school right like so districts are you know range hugely in size but if you’re trying to get at numbers and if you’re trying to get at numbers of kids furthest from opportunity then your urban districts are a good place to start you also obviously have rural districts as well with a lot of underserved young people but it’s a just a different kind of a challenge and I know that you guys have done some work in rural and indigenous communities so you know a little bit about that challenge and then I think what’s happening now with the sort of devolution of I mean states have always held the sort of constitutional authority for education but the interplay between states and the federal government in our current moment is interesting and I think that states are going to have a lot more I mean it’s tricky because we know in terms of policy there’s some barriers but in terms terms of practice I think states are going to have a lot more ability to say hey this is what we are doing this is how we are taking on this challenge and that could set the stage for what the districts do so I think that you know to tackle the problem of change I would say it’s definitely not a school by school we cannot solve anything school by school not happening so I believe school districts because if there’s a reinforcing infrastructure between what’s happening in the school and what the district expects then you’re going to see more traction and then the moment we’re in as as challenging as it is in so many ways I think pres presents an opportunity for states to be innovative and to put a stake in the ground around some of these issues in a way that could also stimulate philanthropic funding. It could stimulate, you know, local policy. But to sort of to your specific question, the community is important.

I need to feel like I’m not alone. I’m not the only one doing this. I need to know if I run into this problem.

There’s someone I can ask that probably has used the same app or the same program that might be able to say how they fixed that problem or how they dealt with this privacy issue or are you getting weird answers or whatever. Like that is really important for the social aspect of learning. And even in the research on the work that we did at Gates, it’s not out yet, but it’s going to suggest that the networking of schools together was actually a driver of change.

And I think that’s going to be very important. And I think that’s very relevant here. Yeah.

This yeah this idea of networking schools together it’s it’s exciting to me because it’s not that we have to build all these networks like there are a lot like like there’s so many there’s so much connectivity yes between schools and like our rural work being the example of like you know how do you reach you know all these like tiny like sometimes 300 student districts and well there are small and rural school associations in not every state but I think most states there are myriad nonprofits that serve small and rural schools, whether that’s state nonprofits or national nonprofits or regional nonprofits who in some cases have spent decades building that network and they know exactly who the nodes are and who the who would you want to have in the room if you really wanted to get something that’s going to get some you know take some momentum out of a meeting. So I think part of our obsession is like how do we build on top of all of that work that’s been done as opposed to try to create something new. Mhm.

What about like within a school cuz like one of the other things that you’ve you’ve done in your career is you’ve actually like I mean you talked about your work at like high-tech high but the the the like the change management that happens within a school like you know the we see a lot of instances where there’s lots of like like leaders at a school who are like we want to use more AI or do AI stuff or you know there’s lots of AI initiatives out there but you know much of the time when we talk to folks you know why like you know why is this sort of slowed down or like you what are your biggest blockers? And it’s they’re just not getting the buy in, you know, from teachers or or the teachers are worried or anxious or like to your point, they’re they hear their administration talking about, you know, you should use more AI to help you save time. And what they’re hearing is we are thinking about AI as a way for us to be able to invest less in our teacher workforce.

And so they’re thinking, is my job now at risk once the AI is able to, you know, replace enough of my time? Yes. Even if the administration is absolutely not thinking that, right?

It’s just like there’s there’s just a lot of anxieties that they have to work through. Do you feel like the change management is is that is that getting enough attention in terms of like the problem statement about AI? Because I I think I guess that this question is coming from a place where I feel like we spend so much time talking about the capabilities and the tools and the challenges with respect to how do we make sure that we have the policies in place so that teachers can safely deploy these tools to do X.

Then there’s this other question of like how do you create a condition for like a culture of innovation, right? Which is more amorphous and less, you know, it’s it’s it’s not a it’s not necessarily like one outcome that you can point to, but it feels feels pretty important if we’re talking about like orienting schools not just for what we have now, but for, you know, whatever the world would look like in two years, let alone like five. I mean, change management is about changing behaviors, right?

And behavior change is a result of learning. So if you want to manage this change then the people that you want to change their behavior need to learn right now how much time do teachers have to actually learn right like the US has one of the lowest investments in teacher time that is not n not instructional time you know they get like maybe one period off to eat you know what I mean like you know they might have some time for meeting whatever you know maybe like my school I know they’re closed on next Tuesday cuz it’s teacher PD day, but that’s one day a quarter. You know what I mean?

Like, so if we want to ch to see that change in behavior, we have to let people learn. And learning is social and learning takes time and learning takes trial and error. And learning goes better if you have someone helping you learn, whether it’s a coach or a peer or a principal or whoever.

So I think the question is, are the structures in place to bring about the kinds of change that we want to see? Is the entity designed to does it have the slack? Does it have the the space and time and expertise that’s needed for people to learn in order to change their behavior?

Is there any space for trial and error? Like what if I learn this thing and I kind of screw it up the first time? Like what’s going to happen to me?

Am I going to get in trouble? Am I going to mess up my kids? Are they going to mess up their grades?

Like what are the consequences of it? Right? Like so I think that there’s a structural problem in the way our schools are set up that makes learning really hard and if we don’t learn, we’re not going to change.

And especially when it comes to using something that is so new, people need practice with it. They need reps. They need to figure I mean it was I mean it’s it’s a different example, but even with project- based learning at high-tech high, we were asking people to teach in a way they’d never been taught.

And so we had to get them to understand and have some experiences of learning through projects. So they were like, "Oh, this is really cool. Okay, cool." You know, and when you ask people, you know, a lot of times when they think back to your most impactful education experience in your whole life, it’s almost always a project, right?

So, like how do we expose people like to they use it in your own life? You you know, like just get them familiar and comfortable with the idea of how these things can actually be helpful. And it’s not about taking away your job or your husband’s job or it’s not making it so you don’t have to cook dinner anymore.

It’s just trying to make it easier for you to cook dinner and know what to cook and have what you need in the kitchen. Like that could help me. Right?

So people need to have I think the experiences that sort of like desensitize them around the fear of it to then think, oh now I see how this can be helpful. Now I see. And sometimes having a peer who is successfully using it can be a great way to see you’re doing it with my kids in our school with our constraints and you’re doing these cool things.

I want to try it. So I just I I think the change management problem is really a learning problem and we need to think about how to create opportunities for learning in a very constrained environment that is not really set up for adult learning very effectively. Yeah.

It’s funny because this is like you you start to get to how interconnected all of this is. Yes. With with I think learning at the center because it’s you know one of one of the things you mentioned is it’s really hard to get you there’s only like one teacher PD day per quarter, right?

And we did work with Prince George’s County and we we like we had funding for three years. We had like you know buyin from the superintendent and like all these folks on leadership. We did massive training with like you know hundreds of teachers over the summer and they were meeting with their team to schedule the next training.

We were like oh we were thinking maybe like August or you know September and they were like oh we were thinking March like March like that’s no no we were like we were like we were thinking maybe every other month if not every month. And they’re like, "No, like we yes, we have these training days, but we have a teacher shortage, which means we have a substitute teacher shortage, which means we have fewer inservice days that we can take, and by the way, we have all this compliance compulsory training that has to get in." Mhm. And so we were fighting initially for like 1 hour out of that training.

And and it led us to this realization that, you know, that working within just sort of the confines of that system wouldn’t be enough. And we had to sort of think about how do we push towards sort of that towards that culture of innovation and and those learning opportunities. And you’re only going to convince the administrators by you know getting some teachers who were who were making the case for the learning themselves as opposed to you know us as outsiders coming in and saying this would be helpful which is what worked and we’ve trained like I think over 3,000 teachers now in PGCPS.

I don’t know if you you’ve ever us met Jeff Riley who was a chancellor formerly the chancellor of the Massachusetts Department of Education and we went down the rabbit hole. We were talking about the misp. Are you familiar with the misp Maryland it was like the most famous standardized test because it was like the one standardized test that students actually look forward to taking because it involved all these group projects and it was had a lot of project based learning.

It was it was measuring critical thinking and like collaboration skills and it ran like I think from like the mid ’90s to 2002 and it was a casualty of No Child Left Behind because it didn’t provide student level data. It only provided at the school level and it was deemed like well that’s just not good enough data for us to use and and so that was just as a site I just think was like fascinating. It’s not that we can’t measure critical thinking or we can’t measure you know like project based learning or like the the durable skills that come from project based learning.

It’s just it requires like a bit of a a change in sort of like aperture. But the other thing he said is that so you were talking about the fact that well I think only 18% of teachers today would recommend the profession to someone else. So down from like the low60s not too long ago.

Yeah. And one thing that Jeff said is well, you know, part of like when I was a teacher, one of the best things of being a teacher was, you know, the fact that we got to learn together and there was so much more of an emphasis on, you know, shared learning and like professional development wasn’t this sort of boring like compliance exercise. It was like truly like it was a chance for us to advance our profession.

And so it it it’s sort of interesting like if we can nail the like like getting learning right ultimately is going to make teaching more fun. Yeah. If it’s truly learning and not just like here’s a tool that we want you to use.

And creating a the ability like creating the the infrastructure to learn is an answer to the question of well what happens if the technology looks completely different in two years because then you lean back on your learning infrastructure that you’ve built as opposed to to update Yeah. Yeah. Yeah.

Exactly. Exactly. I mean I like that some of the work that’s happening like with say your your district you know you’ve listed a list of math or whatever you know you have a curriculum that you write.

How do we help you use this better? How do we help you understand this? How do we help you make it make sense for your students?

How do we help you create, you know, activity, whatever? Like I think that yeah, like when teachers can see that this is something I kind of have to do. So, how do you how do we make it better and give me space to learn that and learn that with other educators?

Like that kind of approach, I feel like that that’s the type of thing that can have some traction. It’s going to take some resources. Do you feel like like philanthropy is talking and thinking about you know like teacher capacity and like professional development and you know communities of practice and like change management and you know all these things.

I is it like at the top of the list of priorities? Is it, you know, like how does it sort of compare to, you know, or is it even part of the conversation around AI? Because they’re obviously investing, I mean, like there no doubt like lots of hunters do invest in those things.

What for me sometimes it’s not always the same person though. You’re sometimes you’re talking to the person on the AI team. Yeah.

And they’re like, "Oh, you should talk to so and so on the like K-12, you know, teacher capacity team." And it’s like, "Aren’t those the same?" Like those are the same to me. But I I how do I answer this question? Diplomatically I would say that I think that sometimes the folks who are proponents of technology and want to see more of that technology in schools are not people who understand how schools actually work.

And so either those functions are very separate, the tech and tech adoption and the and the actual like work of schools and teacher learning or the tech adoption overshadows the there’s kind of almost an assumption if we build it they will learn it and they will come you know and it’s like no it needs to be also attended to. And so I feel like we’re starting to crack that nut a little bit with the like curriculum. Like people are understanding if we want to do highquality curriculum we have to pair it with learning.

It’s not just going to implement itself, right? Like I so I feel like we’re seeing some shifts, but I also see this like retrenchment and it’s showing up like in data that there’s a step back from K-12 investment that either folks are going more into the technology and stepping away from like the actual like work of teachers and students, which is concern in its own right, or they’re going, "Oh, well, we need to actually not just be thinking about schools. We need to be thinking about like housing and healthcare and and and I’m like well yes that’s true right kids don’t live in schools kids live in communities and they need all those things but that doesn’t mean that we should it’s it shouldn’t be a zero- sum game where because we now care about these other things we care about less about sort of like K-12 and what’s happening educationally I feel like that sort of core technology of schools to teacher student content is losing market share in philanthropy right now at a time when actually it is really important that it not and it makes me nervous It makes me really nervous.

I feel like oh sometimes like a Trojan horse because we’re the AI education project and we don’t we’re not here to help you implement AI tools. I’m going to change my tune on that because I think we are getting closer to that but it’s more about we’re here to help you do you know make sure that you are implementing them in a way that is teacher centered and advancing all these things. But most of our work is you know like high quality teaching and learning practices that happen in core subjects and you’re really getting into you know building durable skills and you know career connected learning and the more people dig into our work they’re like this is none of this is terribly new or innovative which by design right like actually we like like our goal is not to go into schools and say we we’re here to revolutionize your district because that is they have lots of disruption happening to them.

Yes. Like that like disruption is not the thing that gets them excited but but I literally have gotten counsel in this they’re like well look I don’t it was this was like a bigger consultancy but you know literally like sort of like verbatim they were like we this is a really tough like your your angle is really tough like systems change and school transformation it’s just losing a lot of people are not excited about it anymore like funders spent you know a long time investing in it feel like a lot of them I feel like they haven’t really seen, you know, a ton of, you know, outcomes from all that investment and it just feels intractable. And I worry about this, like the idea that, well, maybe AI is the silver bullet that will allow us to sort of like skip past all of that and just, you know, use this technology to to to solve all the things.

And I think most smart people like will recognize that it isn’t. But I I suspect that there’s almost like maybe it’s others in the organization or you know I I don’t but maybe is that like a misplaced concern that cuz I know funders are like the folks I’ve talked to are do not really think it’s a silver bullet like I think they’re actually like appropriately skeptical but you go to these conferences and it kind of does feel like it’s you know Yeah. Well, who’s doing the selling and who’s who’s doing the talking, you know, at those in those spaces, right?

Are we talking to educators? Are we talking to young people? I mean, and young people know, you know, I have a young person, right?

I have a 13-year-old. Young people know that this is part of their future and I think they want to understand it and they want to have agency as to whether or not they use it. So, it’s it’s not going to go away.

I just think it’s a tool. Chalkboards were a tool. Laptops are a tool.

They’re a it’s a it’s a tool. And it’s it’s a it’s a set of very powerful tools and very evolutionary tools because like you can we you and I created a tool at the beginning of this call like you know like you can you know it’s not it’s not a static tool right it’s it’s a dynamic tool which is different than than most of the tools that have been introduced that are much more static in nature so it’s exciting but you there is no silver bullet for sure we know this and I think if the more that we see it as a tool versus a solution I think the better off we might be because solutions have been pitched to teachers for a long time and none of them have been solutions. You know what I mean?

It’s a tool that helps you do things better, helps advance what you your goals for your stu, you know, like I just again it’s that user centered view and I think that those users are teachers and students you know I think this is not just an adult game like you said there’s ways in which the kids are probably way out ahead of the teachers on some of this depending on their access depending on their savviness. So how are we thinking about you know ensuring that it is being seen as a tool and not an answer? Mhm.

It’s I have a I have to go. Well, thank you so much for sticking with us.