Teach for All's AI for Collective Leadership - A Conversation with Stephen Jull

Welcome to the Future Learning Design Podcast.

By placing ourselves in the middle of the conversation, it's not just to have the conversation, but is it possible for these billion-dollar companies to pay attention to teachers across a global network? And are they interested in, like, mitigating the risk and maximizing the opportunity within this, probably one of the most important verticals, which is education?

Hi everyone, and welcome back. This is the third episode in this four-part series exploring the ways in which artificial intelligence is impacting the lives of teachers and young people around the world through the lens of the recently announced partnerships with Anthropic. In this episode, I explored Teach for All's thoughtful approach to these big questions with the global head of AI and edtech, Stephen Jewell. How is collective leadership in Teach for All's 63-country context enhanced and extended by the creative use of free frontier AI models? Models, and how are they holding critical questions of equity, access, and data sovereignty as they build communities of educators across the globe as co-architects of AI pedagogies and of the models themselves? Stephen is the Global Head of AI and Educational Technology at Teach For All. Following an early career teaching in remote communities in Canada's far north, Stephen earned his PhD at the University of Cambridge and has spent over 15 years building teams and strategic partnerships to deliver edtech innovations at scale. He was the co-founder of GeoGebra, one one of the world's leading providers of dynamic math education software, and he has supported many young entrepreneurs and high-impact, high-growth startups and scale-ups in roles such as visiting fellow at the University of Cambridge Judge Business School and entrepreneur in residence with Founders at Cambridge Enterprise.

Just super happy to, to be able to talk with you about what Teach For All is doing. If you could just start by just laying out a little bit of the context of Teach For All, because obviously listeners will be aware of who Teach For All are broadly, and they can check out obviously Alex Beard and Wendy Kopp's episodes previously, but what's the AI and, you know, broader kind of edtech engagement vision? You know, how are you thinking about that first before perhaps we then talk about Anthropic specifically.

Yeah, it sounds great. First of all, it's super important to everyone to recognize that I'm a newcomer to the Teach for All community. Lucky staffers who get to be a part of this remarkable project that Wendy started many years ago and started of course with Teach for America on the backs of her undergraduate thesis.

Yeah, really amazing.

Yeah, it is. There are very few people in the world and you can count them, we can name a few both in enterprise, and in social enterprise that can carry an idea through for that long and have it stand up. So it's, it's real rare. And so its thesis has really borne out and it was always on my radar as an educator. So I've been in teaching for a long time since I started as a teacher and worked my way through academic research and building and here I am with, with Wendy. So the context for, you know, AI and Teach for All comes in the context of the core thesis of Teach for All, which is all about this idea of collective leadership and systems change, right? And by attracting some of those who might not otherwise consider teaching or who would consider teaching and give them a unique experience to be placed in close proximity to some of the most important questions and issues arising in their communities, which shapes the way you think about the work, shapes the way you think about your communities, and in fact just forms you as a, as a person, you know, contributing to communities into the world. So, you know, at some point, if with tech changing as, as fast as it has, you know, and I would say that AI, unlike edtech, you know, as a general purpose technology affects everything. Edtech, on the other hand, it doesn't because it's very specific for very specific projects, very much like, you know, the project that I worked on for years. And so by proxy of that, you know, while you might be able to be an organization like Teach for All and the Teach for network of partners around across the world, of which there are 63, while you could probably get away with not including a significant focus on edtech because it's very singular and very focused on either topic or pedagogical or curricular specifics, AI is just so general purpose that it's transforming all organizations. Context is clear that it's one of those things that of course Wendy and the team recognized they just needed to take a view and take a position as a multilateral playing in the space. Its core vision is to retain the kernel of what Teach for All is. Of course, this notion of collective leadership, working across this collaborative network of 60+ countries, bringing all this creativity together. But in doing so, we're bringing AI in, make sure that we are placed in the center of the conversation. And this is where we get into some of the jargon that we're building out. And jargon, I say affectionately, about mitigating the risk and maximizing the opportunity, putting teachers at the center of the conversation. Shifting away from a dependency model on old ways of thinking about edtech and technology to an agency position where teachers are using the tools, being able to work with them and build them and have the confidence to stay, you know, ahead of that curve going forward. So in essence, ensuring that educators that are coming through our program, you know, are not feeling like they're being left behind by what is a technology that's sweeping across communities and schools and education systems worldwide. Yeah.

And of course all the young people are all over it in many, many contexts. So yeah, absolutely. And where in that picture does Anthropic come in? I mean, at what point did you then start thinking of, you know, a large strategic partner like Anthropic to kind of support the work?

First of all, so, I mean, Anthropic was our first choice for a number of reasons, and I'll get into that. But I should just preface by saying that as an organization, and I believe this too as an educator, We are agnostic in terms of the technology. One thing that we recognize, for certainly a notion of collective leadership, and I'd say educators and innovative teachers by their very nature, is they're not really big on being done to. So they're definitely collaborators. It's notional that it's actually, it's part of the kind of the whole ethos of education systems, certainly K-12, of collaborative learning and people coming together to find their way. With that in mind, we have an interest and are engaging with all the frontier model companies as best we can. We're moving those conversations along and we see them as important conversations for our community and for the work that we're thinking about. Anthropic was the natural first choice for a number of reasons. I could start by saying they have a really nice founder story and people can look that up and track that down. There have been some moments in the course of AI development where there's been these inflection points and the founding of Anthropic was at a certain inflection point in the history of AI development. So there is some alignment in terms of how they perceive the work that they're doing, how they present the work that they're doing. Interestingly, of course, in the news that's just come out recently, there's a lot of questions about the consequences of positioning yourself in a certain way. And so it's really another moment where we're starting to see each of these models and the companies and the leaders behind them having to take some clear views about where they want to see themselves in history around the development of the products. Now, that's not a comment necessarily on the other models and their views, but it was just clear from the start that there seemed like a really nice alignment with Anthropic. Also, this is going to sound ironic, but they are a startup, more of a startup than, say, OpenAI or certainly than Google. And Google is the incumbent, the big player in the ecosystem in education anyway. I mean, Microsoft in terms of enterprise, they own whatever it is, 90-plus percentage of enterprise agreements worldwide. They're in that space, but of course they weren't building their own model. But what you've got is with Anthropic, they're basically a scrappy startup. When you want to do something that's new, it's always easier to work with an organization that's just feeling like, yeah, let's just go, let's just get this done, let's try some things out. They were really amenable to the idea of working with us and testing these ideas out around dependency to agency, putting at the center of the conversation. We believe also in this notion of reciprocal learning and reciprocal influence. And that's kind of another key component, which is we are not a market funnel for these models to gain access to new regions or new geographies. But what we are is a partner, a true partner in understanding how the models can be used in different locations effectively and how the context of how these models are built is incredibly important. And that was just something that really resonated with the Anthropic team. They're like, yeah, we can see that. We can work with you. So it was a really nice fit from the start.

Nice. No, that's great. And just to clarify, by that reciprocal, you mean then the use cases, the way it's being used on the ground in all these different contexts might then feed back into the architectures, the models, the way that things evolve in the future?

Yeah, I mean, that's the thesis. And so Wendy and team, we've worked on a couple of core theses to the work around how we're thinking about AI. And that is one of the core components, which is exactly to summarize. So there's a notion of reciprocal influence. And we, by placing ourselves in the middle of the conversation, it's not just to have the conversation, it's, but is it possible for, you know, these billion-dollar companies to pay attention to teachers across a global network? And are they interested in like, you know, mitigating the risk and maximizing the opportunity within this particular vertical, which is education? Probably one of the most important, you know, verticals they need to pay attention to is the future of our communities, our society, our economies. Yeah, so that's really like one of the core kernels of the partnership.

Yeah, great. And I saw, I mean, I like the phrase co-architects that I saw in the press release. I mean, you know, seeing teachers as co-architects, but it's even more, it's deeper than that, just co-architecting the use, but also then co-architecting the actual models going forwards. That takes it to another layer, I think.

Yeah, yeah, I totally agree. And that's kind of the interesting part of this is that it's not kind of like a core driving thesis, but, you know, something I've always believed is that teachers are inherently creative. You know, it's like, it's just a fundamental component of like, if you go into education, you either are creative before you arrive or you are when you leave. You just have to be because, you know, I mean, you work in the field and, you know, in order to kind of like navigate just the complexity of a classroom, which is absolute complexity. Like this is the— people worry about teachers disappearing. They just won't because classrooms are so incredibly complicated that, you know, I'm not sure there is a model. Maybe when quantum, you know, computing Finally get that back, it picks up. And I think Cambridge has bought their first, you know, quantum computer. It's gone in the physics lab over here. When maybe then they'll be able to model the complexity of a classroom. Like, honestly, good luck with that.

Yeah.

Yeah. Good luck with that. Exactly. So the opportunity for Teach for All to, you know, be taken seriously here is, is really true. And teachers are naturally creative and exploring these tools. And by proxy of doing that, it's been amazing to watch our colleagues at Anthropic and some of the other models who are starting to work with. Kind of go, oh, right, okay. So people are developing pedagogical solutions, identifying curriculum, doing the work that edtech used to do, but they're doing it completely independently and even with the free models. And that's kind of powerful. Yeah.

Nice. Oh, great. And then, I mean, just getting into a bit more of the specifics, correct me if I'm wrong, but as I've understood, you've got this AI Literacy and Creator Collective model, which has different parts to it.— that as the way that the programs are perhaps kind of rolling out slightly differently based on how much a teacher knows about, you know, how much exposure they've had already, where they're at in their own journey. Yeah, perhaps could you just talk a little bit about those different elements of the LCC model?

Yeah, thanks for that. And that was also kind of like early stage with the team. So we're all quite new to this and we tossed around like, well, how are we going to manage this relationship? You know, how are we going to like lean into the thesis, you know, that we kind of set out here that we were trying to pose here on how AI can mix in with collective leadership. And just like you do with any kind of agile development, it was a very startup-y process. The selection of the words, I know this may sound a little bit pedantic, but it really matters. Of course, AI, you'll notice that edtech is very explicitly not there, even though we are the AI and edtech initiative that's building this work. And we still are paying attention to edtech, so just so that everyone understands that edtech is still an important component. It is an application of technology in the context of education, So it lives as an idea and as a vertical. But in this particular group, we were just going to focus specifically on AI. So it's the AI, and of course literacy, fundamental, and creator collective. So the idea of going from literacy to creator was really kind of intentional as well too, because, you know, there's no upper limit to being a creative. There's no ceiling. So from coming on board and just finding your feet, exploring it, building your confidence, being around other people. Who are also just sharing what they're working on at all different levels. There's a real kind of humility in the whole group and how we collaborate, but we can also see people rapidly moving towards the creator kind of journey, and that's really the natural direction of travel. And it comes back to this notion that teachers are naturally creatives, so they want to build, they want to create, they want to just keep moving. And of course, collective is, you know, back to the core vision of, you know, Teach for All, which is collective leadership. So by pulling all those things together, it allowed us to kind of create both the semantic framework, but also kind of the programmatic components that enable us to build out the AI Literacy and Creator Collective as a community.

Nice. And how much of that is leveraging existing Anthropic? So again, correct me if I'm wrong, but what I understood is the AI fluency learning part was kind of an Anthropic pre-existing that supports the literacy part perhaps. And then you've built, then building out communities off the back of that and labs off the back of that.

Yeah, it's a good question. So I mean, fluency is a word that they use. It's not a word that we use. Okay. But that's not to say fluency— I like fluency too, but fluency implies like, I would, I haven't actually had this conversation with my perfect colleagues, but fluency implies like a level of achievement. Right. And so, and when you think of fluency, you think of language, right? You can think of math fluency too, but the first thing people think about is like fluency in a language, which is not your first language. That's scary. Right. So becoming fluent in something is like, yeah, Yeah. So there's the literacy, you know, is, is another, I think it's a more inclusive and I don't mean inclusive just to be inclusive, but it's, I think it spans the spectrum. But yeah. So, but to your question, while, you know, certainly they have a view on, and so the other models on how they engage with, you know, through fluency curriculum, some is called training curriculum, like there's various ways of framing it. Okay. We think about it as literacy. So are we consuming, you know, their core curriculum around how they want educators to take up these tools. The short version, Tim, is, is no. You know, so it's not that we don't appreciate the curricular components and it's not that we don't think there's amazing opportunities out there to, you know, pursue some of the off-the-shelf courses that are available across all the models. But we really took the view, and Anthropic was very open to this, that we're going to create bespoke, open-ended, very relational, but very tech-baked-in live learning sessions. And then asynchronous sessions in our WhatsApp community, which has almost 600 people in it. You know, it's like about 50% of the total registration is very active in this very lively community, all organized around individual collectives by language, by region, by topic, you know.

Nice.

Yeah. It's a very manageable space and, and very, very lively. And so, you know, they were just also very open to this idea of like, okay, we get that. You know, even though we're by proxy, like when you think about the marketing spend of a big organization like that, when they produce content to ship, you know, to a region or a geography. They think in tens and hundreds of thousands that are going to be viewing their content. I mean, with us, we're big if you take our global network, but our early nascent emerging AI literacy and credit collective is a small community of 1,000+ teachers, educators, alumni. And they were very open to the idea of like, okay, we get it. Let's create this space for educators to learn together with us. And so what did that actually look like? It would look like on many occasions in most of our live learning sessions, which are really well attended, We would have toe-to-toe in the room, you know, people from our community with colleagues from Anthropic or other special guests having conversations, doing live demos. We don't use a lot of decks, so it's a lot of live building. And that's kind of how we collaborated with them. Yeah.

Nice. And then, I mean, I quite like the idea of this all happening on WhatsApp, you know, which is clearly not an analog platform, but it's like, it's further down the stack towards analog. You know, you've got self-organizing communities happening all over the world on WhatsApp chats. Like there's something really much more real about that, right?

It's really good because, you know, we had that conversation as well too, as you can imagine, like, yeah, which platform were we going to look at? Of course, we looked at all the usual suspects, you know, and we knew we had to build somewhere that because we didn't, you know, we're obviously like operating as a startup on a shoestring. WhatsApp's what's in people's pockets, but it also has opportunity down the road. We see it as a really powerful platform, as many do, in both business and education. And it actually serves education in the global majority in a really big way, certainly in English literacy and mathematics and other places. You've seen these incredible founding startups that are in that space. But we see a great opportunity also to plug some of what the powers of WhatsApp allow into future ways of thinking about how we make this a sustainable and growing community over time. So yeah, it's what people have, so go where people are.

Exactly. Yeah. And I mean, I lived in Tanzania for a while and I remember just seeing just these incredible creative ways that WhatsApp was being used, way more advanced than now living in France where I live or other places, just phenomenal. The kind of that, and this wasn't just in education, but just the creativity to adapt and use what's available. Quite amazing.

Yeah, it's really true. And there's a tone that changes on WhatsApp than other platforms too. There's something about the easiness of it. We're all so familiar with the cadence and the length of the— So I think because it's so universal, it actually just allows people just to relax a little bit as well too. So yeah, it served its purpose.

It's like a leveler somehow across 63 countries.

Absolutely. Yeah. There'll be billionaires on there. There'll be people just Starting their first year, if you like, you know, you're grade 12 heading to university.

Yeah, yeah, yeah. Interesting. So then just again, just to kind of ground it a bit, are there any examples of the way this shows up in different contexts that you might just give to kind of make it real for people? I read a few of the kind of the examples through the work that was shared, but you know, it's nice for people to hear like, how are these teachers in Liberia or wherever being really creative and using this stuff?

Yeah, I mean, it sounds like I'm skirting it and I will say something specific as well to you, but there are so many examples that happen every single day that it's just, I wake up, it's a bit like back in the day with GeoGebra, you know, like when we were just building the platform, the best part of my day, you know, was waking up and seeing the new GeoGebra applets as we called them that were publicly available on the platform and there'd be thousands of them, right? And it was just like, oh my goodness, I didn't think people Who would imagine building something like that? It was just amazing. Like, it was just so, so cool. So every single day I started my day like that. It's like this now. So like I wake up, when the first app's open, I go to the AALCC. I look at all the groups and there's just like tons of stuff in there. And I start my day by getting inspired by the work that people are doing. And so it's everything from pedagogical innovation to talking about safety, to working on curriculum, applied work, you know, in a specific domain. I'll just mention climate, you know, because that often comes up in, you know, emerging, you know, locations where we need to be thinking like sensibly about both the technology and the opportunity. So it spans curriculum, pedagogy, safety, policy, you know, just the works, right? And communication, operational innovation. Actually, let me, let me go to that then, because this actually happened just the last like 48 hours. So the community spontaneously decides, first of all, they collaborate together, like, you know, as small groups and in these collectives and, you know, lots of sharing going on. And then just recently in the Claude lab environment, which is like a 12-month tenure, you can apply to join. We're very big on the free models because it passes our equity, scale, sustainability, impact test. And that's really important to us is that we don't try to build bubbles that are dependent upon fee-based products. And that's another great reason to be thinking about the free models. But we also thought it might be interesting for some of our members who are heading towards that builder kind of end of things, like moving from creator to builder, to have a space to really push the models really hard. And for Claude Lab, it's a new program that just started and it's got about 125 people in it right now and growing and it's very lively and there's criteria to be in it, like innovating and sharing. And so it's, it's very full of like innovation. Anyway, recently one of our CawdLab members decided that they were just having such great fun with, with coworking. And that's all the rage, of course, and everyone loves coworking. So one of our members just said, hey guys, I'm loving this. How about I just, I'm really digging into this right now. I've got the time to do it. I'll hold a little mini workshop. Posted in the WhatsApp community, people signed up, spontaneous, you know, 1.5-hour event. The next day I wake up, so that was brilliant. I jumped in for part of it, you know, between meetings, et cetera. It was wonderful. It was fantastic. The next day I jump on and someone said, that was amazing. I love that. I just did a review of, we're talking hundreds of documents at the organization that this person happened to work at and it wasn't in education, but it was in social enterprise that enabled her to do the work, you know, in a way that she might not have otherwise considered using Cowork. So that level of like, and not, this is not even the same location. So like just, just by proxy of participating in this community, the level of innovation just kind of like spreads and it just creates this like virtuous cycle of, of like innovation. Again, literacy to creator, it just keeps that flywheel going.

Love that. Yeah. I mean, it's super interesting, isn't it? Because it's like, of course you're gathering around the AI broadly frame question, opportunity, whatever. But there's just something really dynamic about the learning. It's like, it's a very alive community of practice, right? That you're just kind of, you're seeing these people just super excited and enthused about something and sharing and then offering and learning together. And it's, yeah, that, that's like, as a human dynamic, that's a pretty cool thing, right?

Yeah, 100%. I mean, that's, that's why we're in education, you know, that's why, you know, do the work that you do as well too. It comes back to the various core questions about will AI replace teachers? Will it completely transform education and make it unrecognizable? To your point, fundamentally we are a collaborative species. We like being around each other. We trade our ideas. We might get these really, really powerful tools. At one point it was fire and then it was other things and now it's AI and subject to us ensuring that we use it responsibly and to begin put some effective guardrails around it to make sure that it presents more value than hindrance, it, I think, will just create that opportunity for people to be even more creative. And I think educators are naturally inclined in this space. They just love to create and innovate. And I'm like you, it's kind of a joy to work in this space when you're around people who are just innovating continuously.

No, that's great. I love it. I mean, it's, yeah, we off mic, we had that productivity conversation and there is something that You've, with the automation and the, this kind of obsession with increased productivity, like, yeah, it's great that you can get it to create all this stuff, but actually the real joy and the meaning and the purpose that people get is it actually in those learning moments and communities. It's not necessarily about, yeah, I just, I got it to summarize 300 documents and produce me this thing that, yeah. So it's, I don't know, it's, it's just to reflect, it just feels we don't wanna lose that part of it, I guess. It's part of what—

Yeah, this might be too trite, but you know, there's a reason why show and tell was everyone's favorite subject when you're like being there. Because yeah, that's it. There's no joy in creating alone. Like you create and you have to go share it with somebody, right? So that's what makes it all good.

And I mean, you know, this is all great, right? You know, I'm conscious it's not all a happy, fluffy world of all these joyous learning. You know, there are some pretty serious questions around a lot of these things, and particularly in many of the jurisdictions that Teachable is working in. So I guess I do want to just touch on that if we can, just like, how are you holding those questions of the implication of these big AI companies in the more ethical questions of obviously how the models are trained and all of that stuff, exploitation, and then also, as you said earlier, kind of ensuring that it isn't just a harvesting of data for increased global inequality patterns to just rocket fuel those, but actually something else, a different kind of dynamic that I'm pretty sure Teach for All stands for pretty strongly, right?

Yeah. Well, there's a ton of things in that question like that, none of it which I would disagree with in terms of risk. So we could talk about all those moving parts which all lead up to equity issues and increasing gaps and issues of privacy and data sovereignty. There's so many conversations. They're very, very front of mind. And I'm happy to drill down into some of those too and how we're actually handling each of those questions. And no one's ever perfect in these domains, but we should be paying very closely attention to it. And that comes back to this core notion of like, what's the core vision and value of Teach for All? So we have the benefit of falling back on this fundamental principle, which allows us to always steer in our direction, in the right direction. But so here, you know, at the top level, you know, we took a view that anytime there's a huge potential shift coming, or whether that's a policy shift or change in way of thinking, and in this case, of course, a big general-purpose technology, then you have two options really. One is you can turn away or you can turn towards it. And so we have just taken the view that we're going to turn towards it. And so we're walking directly towards the questions and the opportunities and the problems, and we're dealing with them at each point along the way. So it sounds like a metaphor, but it's actually kind of in practice is how we're approaching it. Yeah, at each step of the way. So let me stop there and see if that triggers, if we dig—

No, no. I mean, I think you're exactly right. It feels, it's like a tsunami that's like you basically just can't avoid it. I mean, it, so even the choice to turn away is actually increasingly difficult to do anyway. And so that's partly why I was interested in the work that Anthropic is doing within these different situations because of course there are lots of challenges, but how might we as a community learn to turn towards it and figure this stuff out? And I think it's really important.

Yeah, I totally agree. There's something else I would add to it as well too, which I kind of been thinking about probably a little over a year ago when it felt like there were waves of development and the waves interestingly, they were really fast. There's that period of time where every week just was like, oh yeah. And the change I think is actually kind of slowing slightly, kind of a natural kind of cadence, or perhaps we're just becoming acclimatized to it and it feels less like the changes are as rapid. But I was at that point, you know, in talking with teachers at events, et cetera, because there was a lot of anxiety like 12, 18 months ago, right? And the anxiety level is coming down. Teachers are actually becoming very, very confident. It's another very interesting thing that I'm seeing is like of the sectors worldwide, if you had to find the vertical or the kind of the sector that is probably most confident with these tools or exploring the most, educators.

Oh, interesting.

Right. It's a natural landscape. And so I think they're leading the charge. And that's a great thing because that's going to help the models move in the right direction because they're the big power users. But, sorry, back to where I was. Even though we need to turn towards it and face the opportunity and problems, mitigate risk, maximize the opportunity, you don't have to jump in straight away. I say this to educators as well too. It's okay if you just— it's fine to take a pause if you're not using them yet, which people are still kind of exploring. It's okay. The change will keep coming. The right point to start working with the models is the time when it feels right for you. It sounds like a throwaway comment, like a very teacher-y thing. It's like, when you're ready to learn, you will learn. But it's a truism too, right? But the good part about it, of course, is that when it does change, it changes so much that if you started 6 months from now, you would be just about where everybody else is at the same time because it's all new, you know, each time. So that's kind of the joy of this technology too, is if you're not there yet, you don't feel like you're right at the bleeding edge, that's okay because it's changing so rapidly that the time that you jump in is the time that it'll be that time and you will catch up. And it'll be fine.

No, it's interesting, isn't it? And just on that equity point, how are you addressing that question in some of the really under-resourced communities that you're working in?

Yeah. So there's a couple of things that, like, all of these questions are so enormous.

Yeah. Sorry about that.

No, no, that's okay. Because, yeah, this is a big change, right? So these are good questions. First of all, in practical terms, the work that we're doing across the whole network of 63 countries and many, many thousands of educators, right? Which have impact on millions and millions of students. We're working with the free models. That's our premise. When we're out there building, we're building with the off-the-shelf free models. Free is not always free. Free comes with— obviously there's data sharing, etc., that comes with this, of course. We take a view on this and we turn towards that and we have those conversations as well too. In order to make this meet our equity, scale, sustainability, and impact test, these have to be models that everyone can access. In terms of just straight up equity to accessing the opportunity, if we're starting with the free models in the first instance, then it's a device connectivity and data problem. And we're also looking at those problems too. Those are real problems because there's, I think, nearly 2 billion people who still don't have regular access to the internet. And that's a known stat worldwide. So if we can resolve those other problems with strategic partners and/or some of the work that we do as well too here, then by working with the free models and resolving device data and connectivity questions, then we're at least in an equity position. We can all have access to the knowledge economy, the knowledge opportunity that AI presents itself. The bigger question, and there are many, but one of the ones that's really front of mind for me is even though I'm just one person in this organization and in the world of this conversation, I strongly believe that the free models, if they are perceived by governments, and we were citing one government that's just signed a recent deal and there are many that are signing, with these big frontier models. If we're perceiving this as a government, as society, as essential for our economic sovereignty, our economic opportunity going forward, and that applies to everything, then you can't exclude some people because they can't afford the same models. I see that there's both a responsibility and an interest on the side of the companies developing the models to make sure the free models are just as robust, just as powerful. They may not be as fast, maybe they can be constrained in other ways because I recognize these are businesses and that they have to operate, they've got constraints themselves. But I think there is, in terms of resolving the equity issue around access to knowledge, which is fundamentally what this is, it's an access to knowledge opportunity, then if we can solve the problem of free models not being less able, then I think we're pretty close. I think we're getting to that point where, at least in terms of K-12 education, and the opportunity that this presents, I think we're going to get pretty close to equity. And so that's one of the kind of core—

Yeah, no, I mean, I recognize they're huge questions and we haven't really got time to go into the detail, but no, I do appreciate how I can totally imagine having engaged with so many of your colleagues that, you know, how thoughtfully you are all engaging with these questions. So no, I just, I really appreciate it. And thanks, Tim Logan. Really good to chat about this stuff.

Yeah. Amazing. Well, thanks, Tim Logan. Yeah, it was great fun. It's a nice way to spend my evening and, uh—

We hope you've enjoyed this episode. Please feel free to continue the dialogues with our guests, with us at Good Impact Labs, on our blog or on social media, or within your own networks.