AI in Schools: Should It Be Banned? | MagicSchool AI Founder Adeel Khan

[00:00:00] Adeel, this is going to be a really interesting one because AI is obviously a big topic but we're going to talk about AI from a perspective that I think is a little different than many of the AI conversations because we're going to talk about it from the perspective of MagicSchool. And let's start by, and we'll get around to introducing you and who you are and what MagicSchool is, but I want to first talk about debunking a few things that I think are out there, or at least get your opinions. I assume we'll debunk them.

[00:00:28] And I want to hit two or three of them in whatever sequence you want. One I want to talk about how AI, whether or not, and if so, how AI should be used with kids as they're going through school. You know, is it something you should treat like social media and say that doesn't belong in a school and we're going to make you learn the old-fashioned way?

[00:00:50] Or do you think AI should be a tool for teaching kids at a young age and if so, why and how do we do that? Which obviously plays directly into MagicSchool. I also want to talk about the effect on teachers because I can imagine that if you're a teacher who's been teaching for 10, 20 years and now all of a sudden AI hits, the level of challenges that I know I would have if I was a teacher and knowing what to do with it would be high.

[00:01:19] And I want to get your take on that. And then I want to talk about AI and maybe, and again we could intermix these together. Is AI a force of good when it comes to creating opportunities for that next generation of all different socioeconomic circumstances?

[00:01:36] Or is it something that should be viewed as a threat to the opportunities that will be available for kind of kids who are now coming up through kind of the K-12 school system? And if so, what should we do about that? So a lot of meaty topics there and we're going to hit them all.

[00:01:58] I'm looking forward to it. I like that this is like an expansive conversation. It can go a little bit long. I feel like I have a pretty unique perspective. Yeah, let's talk about who you are now and then we'll get into all these crazy topics. Yeah, I think that there are probably few people in the world who have the experiences I've had which have been, I was a classroom teacher. Yes, you were. I founded a public high school here in Denver, so I was a principal there. I led at a district level. I managed principals.

[00:02:26] And now I'm a founder of an AI company that serves K-12 education. Let's just pause for a second there because you said it already, but I want to make sure we just put a fine point on that. You were a teacher. You were a principal. You founded a school and now you've launched an AI startup company focused on education. I mean, who else does that? I mean, you might be one in a kind of unicorn of a different type than thinking about, although you're probably on path to be a unicorn of that type as well.

[00:02:53] But unicorn as an individual for what you have done leading up to the launch of this company. That's just incredible. Well, thanks for the kind words. The more I think about the conversation that's happening at a national level with generative AI in K-12 schools, I'm not one to seek the spotlight. But I am someone who think I can offer a valuable perspective in this debate because of those experiences that are like earned experiences.

[00:03:20] So, you know, when I talk about the way that a school operates, it's from like a place of like I've seen the way a school operates. I know teachers. I was a teacher. I can interpret the technology. It's both the thing that I think allowed for Magic School to scale the way that it has is that when generative AI came around, I could see so clearly how it could make positive impact in schools and also the associated risks and create something of what I think is like a brand that is trusted in K-12. We partner here with Boulder Valley School District. They're one of our first partners.

[00:03:49] They love Magic School and they think of Magic School as this thing that is both helping teachers do their daily work, but also preparing students for the future. So, again, while I don't seek the spotlight, you know, as I've lost some hair, I feel like I have some valuable insights to share with folks who might be interested in what the AI and education conversation is. And you know what? Like just about everyone is affected by this.

[00:04:15] I talk to folks who have never worked in education and they ask me like, what do you think about my child's education? What kinds of jobs might be around in 10 years? What kinds of skills should they be developing in schools? And, of course, the questions that you brought up at the start of the podcast, which are how might schools introduce this technology? Should they? Should they not? There's so much in the news right now around this topic.

[00:04:43] LAUSD banned generative AI for students in schools. New York City public schools did too, the two largest systems in the country. And, you know, to study and find out, you know, what they should do. I don't think that's going to stand very long. To me, that just seems the exact opposite of what you should do. I mean, to not take advantage of the fact, well, one, first to not fully reconcile the fact that AI is here for good. I mean, there's no going back.

[00:05:13] It would be like saying, you know, this Internet thing is pretty threatening. Let's put off teaching our kids about the Internet for a few years. That would have been crazy to not introduce and leverage Internet in education as soon as you could. And the same with AI. And so the thought of school districts reaching a conclusion that let's not allow or let's be very careful about how we allow AI to be used in a classroom could be very dangerous and could set a lot of kids back relative to their peer groups for a long time to come.

[00:05:43] Because there's no catching up what you learn when you're in first grade, second grade, third grade, fourth grade. I mean, those are such formative years. And if you have a group of students who are benefiting from kind of the leaning in use of AI, introducing that thoughtfully and effectively to kids early on, and you compare that to a group who is sheltered from that,

[00:06:05] I got to believe there's going to be a permanent gap in the life journeys of those two populations to the benefit of those who, even if there were mistakes being made, who are using AI often and early in their childhood experiences. Yeah, I think that there's a, the idea to ban generative AI in schools at any level, I think is short-sighted. That's sort of what that means.

[00:06:36] Generative AI, I mean, I know what generative AI, I assume most people listening to this know what generative AI means in general. But when we say ban generative AI, that could be a very narrow term or it could be a very broad term. But it strikes me as broad in that context, broad by banning generative AI. A lot of things would be banned from schools under the banner of it's only generative AI. It's not AI, it's generative AI.

[00:07:02] Well, my guess is that means, you know, whole categories of AI that the wrestlers are using every single day. Yeah, I mean, even in the sense that like, if you ban generative AI today in a K-12 school, and right now, the current policy as it stands in New York City, which I believe will change. And remember, this is the biggest system that serves children in the United States, the largest by student population, is that you'll get two 45-minute AI literacy sessions at some point in your high school career. And that is the extent with which they've committed to using AI with students.

[00:07:30] You said they serve their students. That doesn't sound like servicing. It doesn't sound like a disservice. It doesn't sound like servicing kids to me either. And again, I think they will change this. That's my hope at least. But if you think about what it would mean to ban generative AI for kids, again, your point is well made. You can't really do it. If you Google search something, the first thing you see is a generative AI chatbot that tells you an answer to the question from AI that you can continue the conversation in. So it is already Gen AI, a chatbot. Gemini shows up if you do a Google search.

[00:08:00] So it's kind of everywhere. The thing that bothers me the most about these policies is that it stands to – I guess it suggests that AI is only used on student school devices and the only place it exists for a student is in a school building.

[00:08:17] But the only thing you've done with a generative AI ban is you banned it in the time and space where the most responsible adults and sometimes in some children's lives can help them understand this technology, use it for the good reasons to use it, protect from the reasons it can be harmful like passing off your thinking, learning loss, those kinds of things, the cognitive offloading that it can do.

[00:08:42] If you're not talking about that, then you're just leaving kids to their own devices at home, which will have this technology in every place that they already use tech. But they won't have it at home in equal measures. They won't, yeah. Certain people will have access to their own, the latest models with highly powerful computers and a lot of insight and direction around them that enables them to use it. And others will go to households that maybe have a computer, maybe they don't, maybe it's an older computer.

[00:09:12] Yep. Maybe the access to the generative AI models might be nothing more than a search engine that has AI built into it versus a true model. So if, you know, the ones who are, I mean, you talk about like a divide, we often, you know, talk about the broadband divide, which was a very real thing, but it played out more slowly over time. And maybe we were slow as a society to fully understand that and find a way to reconcile it. But now we have this thing that's called AI that could create a gigantic AI divide.

[00:09:41] And it's a lot easier one to avoid than the other one because you don't have to build fiber lines to really hard to get to areas to give broadband to broad populations. But if you choose as a political leadership, you know, structure, if you choose to constrain the exposure to AI that children will get at school, you're just going to create a divide that could be, you know, even if it's only for five years. Yeah.

[00:10:10] While the kids who got affected on those five years live with that for the rest of their lives. Yeah. You know, a lot of the folks who are implementing these bands, the architects of these things, are folks who also say that they value educational equity, which is the whole reason I got into this work. I did Teach for America. That's the mission of my life and my career. And when you talk about a digital divide, that is not equity. It is not equity. When I talked to a school board member, I was in Florida recently, and the school board member is a lawyer.

[00:10:38] She didn't have an educational background, but she ran for school board because her kids went to the district. And she wanted them to get a greater education, thought she could make an impact. And the conversation I had with her, she shared, you know, I actually asked her, why are you talking to me? How did you get to know Magic School? And she said, well, and she called the meeting. She'd heard of us and reached out. And she said, well, you know, I'd heard about Magic School and, like, the idea of, like, responsible AI in schools. And as a lawyer, I'm starting to use generative AI as a knowledge worker in my daily work.

[00:11:08] My kids are going to the school district. So I know they're going to need to be equipped with these skills when they get into the workforce. I'm using it more and more as a lawyer. Well, I think lawyers are going to be the last ones to the table. Oh, my gosh. You work with lawyers on AI, and they just don't trust it. They don't want to use it. It's too easy to get the work done. And they want to give you reasons not to use it. And I think the legal profession doesn't know what they got coming at them either.

[00:11:35] And they got their Harvey, but the rest of us can't use their Harvey kind of craziness. But I think it's a – and you're right. It's a skeptical group. And even she was seeing these title shifts. And she's like, well, I know that my kids are going to need this. A lot of families who have parents who are more well-to-do, they have professional work like a lawyer, a doctor, they know these things are happening. You work in tech. They know these things are happening. So they're advocating for their kids to have exposure to this and learn how to use it responsibly.

[00:12:03] And she was like, I called up our district leader of technology, and I asked them, you know, what are we doing with AI? And they had a really scattered answer. They didn't really know. They said some teachers are using it. Some teachers weren't. So that's why I'm here. And I do want to get into that more too because, again, I definitely can empathize with what this means as an everyday teacher and how daunting this must be. But I want to talk about one other thing first, and you're probably familiar with this. This isn't super novel.

[00:12:31] But there was a study done some time ago about the correlation with those who make the National Hockey League and what month of the year they were born. Now, it turns out that, you know, as we know, looking backwards in time for certain, a disproportional amount of Canadians make NHL than Americans or anyone else. Less so now, but certainly, you know, dominant.

[00:12:56] And what they found is that, you know, certain months of the year were highly correlated to who made the NHL. So they went deeper and said, well, why is that? That seems pretty random. Like, why would being born in June versus being born in October matter all that much in the overall scheme of things? And the answer was it mattered a lot because those who were born very late in the cutoff means that they were the older kids when they were six years old playing hockey, five years old, seven years playing hockey.

[00:13:24] So they would get the most playing time. They'd be at the, you know, at the higher league categories. And then that would carry forward the next year and next year. So those who were, you know, born a month later but was on the other side of the cutoff were always a year behind. And it had a cumulative effect that they never caught up. Yeah. Which in hockey you'd be like, I guess that makes sense, but it's still a bit surprising.

[00:13:48] You'd think that as the kids grew up that would become lesser and lesser effect or almost like noise by the time you looked at the NHL. But I can imagine when it comes to AI, that effect will be there. But it'll be there in much more profound ways. Those kids who are not because of exactly what month you were born, but if you were in a kindergarten, a first grade, a second grade, a third grade, a fourth grade, that is leaning heavily into AI, you know, and you compare that to those kids who were born in the same month but a different school district,

[00:14:18] different state perhaps, or even just in a circumstance where they were in a teaching curriculum that just was less AI forward, that the gap would be pretty big in that first year. It would be much bigger in the second year. And by the time three or four or five or six years pass, the gigantic difference in the capabilities of those two populations will never get made up. It won't get made up in college, post-college. It'll be permanent.

[00:14:45] It kind of seemed like that would be the case. Yeah. I didn't know that this conversation would turn into this one about equity and privilege versus underprivileged, but that is very much what will likely play out. Oh, yeah. Yeah, you're right. So folks with means, they do things like they know that. They know that hockey story, but they also know it for like cognitive development.

[00:15:11] Like, you know, they know that if I hold my kid back in child care one more year and I put them on the later end of like the older end of their grade level, they'll perform better. That same thing is true for kids who get tracked to the like, you know, IB and AP courses, honors courses. So families who are well-to-do will often just hold their kid back an extra year, especially if they're on the early end. I was on the early end. I was like a 17-year-old high school college freshman. My parents wouldn't have had the means to put me in child care for another year and then let me be one of the kids who was ahead.

[00:15:41] But my brother has done this. He's decided that I'm going to hold my kid back one more year and keep him in child care and has the means to do that. But these are these invisible advantages. That would have been a stigma to be held back a year. No one would have done it for the benefit. Yeah, but you do it when they're really young and no one knows, right? So it's like, you know, child care kind of thing. But nonetheless, you know, you do that. These are invisible advantages that people with privilege have. They know those kinds of facts and they do it.

[00:16:05] Similarly with AI, folks with privilege, you know, our first partners for Magic School, a lot of them were like private schools. They were like, hey, like we know this is coming. We want to give our kids access to these resources. We want to teach them how to use it responsibly.

[00:16:18] Now think about those private schools that are doing that with their kids and giving them these builder experiences where they're building their first apps, their first websites, their first presentations using Gen AI as a tool versus a kid who's arriving on campus without saying that this was a terrible thing and something you couldn't use. They're going to have the same assignments when they get to college. And one's going to be vastly more equipped to outperform the other. I mean, that's happening today, right? Like that's happening in the workplace.

[00:16:48] This is going to happen real quick. The broadband divide happened slowly over time and was constrained by the build out of broadband networks or even the conversion of dial up to broadband happened very gradually over time. This is like you said, this is happening immediately. It's already happening. Yes. There's already a gap between folks who get experiences with generative AI in school and when they get to college and not. That exists. And we see it in the workplace.

[00:17:15] Like there are folks who are really AI enabled and can do a lot more work than other folks and can amplify their impact at their organizations with the use of generative AI and folks who are still learning it and they're falling behind. There's also, you know, there's invisible AI literacy things that happen if you don't get taught how to use it responsibly by a teacher or someone who's enabling. You know, companies are hiring enablement folks to enable their workforce to become literate with generative AI.

[00:17:43] But I often use this example when we talk about kids and even in college, the gap that exists. Like, you know, there's so many kids who are graduating college right now who are applying to jobs. They're doing their cover letters and they're using AI and they're using it the wrong way. And they're just doing a generic cover letter with AI that looks like AI and sounds like AI and doesn't tell their authentic story. And they've sent out a thousand of them. They're like, this cover letter is better than anything. This is easy. Yeah, this is the best cover letter I've ever written. I couldn't write a cover letter this good.

[00:18:11] And they're wondering why they're not getting calls back from the employers. It's like, well, if you do a generic AI cover letter that doesn't have your authentic voice in it, you know, you're just going to get – You know who's reading your genetic cover letter? AI. And it's detecting – I can't figure out it's a generic cover letter. You're kidding yourself. That's right. It's just being passed over right away. So there's also this gap of like you have to be able to use it with nuance and thoughtfulness and like, you know, people get turned off when they see – It's a learned skill. It's a learned skill, yeah. It's an art as well.

[00:18:39] I mean, I see it every single day in the workplace. You know, we've got – I like using the word humans now. I don't know if that's proper or improper. But, you know, humans' influence on the output of AI is gigantic. Those who know what they're doing with it and leverage it to its fullest, they are so much more valuable now than they were before and so much valuable to their peers who are also using the same AI tools, but they're doing like what you just described.

[00:19:07] It's kind of like, oh, that was an easy assignment. I'll spend 15 minutes on it. You know, say I got the work done, shoot it up and see what happens. And, I mean, you could see that in a nanosecond. Yeah. Very little thought or effort went into that. And the person who did it, you know, all they did was create more work because you got to go over and redo it anyways. They might as well not even lift their finger to do it. And the graph between those employees shows up so quickly in the workplace.

[00:19:33] And certainly anyone who's doing college admissions is – even before a human lays their eyes on that – is going to have filtered all those out. Yeah, and isn't that what school is for? It's to make those mistakes. It's so that you can put out something with AI and your teacher is like, I knew this was written with AI. Like, you know, if you're going to use it, this is the way to use it. You know, preserve your thinking. Make sure that you're reviewing in every word you believe in before you submit something. You know, like here's the way to use it responsibly and like let the kid make the mistake that when they do that, right?

[00:20:02] Do it in a place where it's low stakes, not when they're applying for their first job. Or they're in the workplace and their boss is like, you know, rolling their eyes at the report that they wrote because it's very clearly 20 pages of AI slop. Like that's what schools need to be doing. AI slop. I got to use that word. That's AI slop. Get it. That's AI slop. Fix it. Yeah. I'll become even more popular in the workplace. And the folks who are listening to this, I imagine, are working in places where sometimes they get AI slop.

[00:20:26] They just get a, like, you know, overly long thing that was written that has, you know, very clear indication that there is very little thought, to your point, that was put into this. And so my argument is that to understand how to use this well, you have to use it and become more skilled at using it. And get feedback. Yes. Sometimes positive feedback, sometimes very constructive feedback. Absolutely. It doesn't just happen overnight. You don't just like turn it on and now you can use it and all of a sudden you use it well.

[00:20:55] So we're going to come back to these themes throughout the conversation, but I want to introduce Magic School at this point. Like, what is Magic School? What role does Magic School play?

[00:21:04] Because when we layer that onto the conversation, now we're going to talk about how the effectiveness of using AI in education will be enhanced, be further enabled by, you know, AI companies building AI products that help provide the means to teach education. So Magic School, what's Magic School all about? So Magic School is generative AI for K-12 schools. It is like the AI co-pilot for teachers.

[00:21:33] So it helps them with their daily work. Things like building rubrics, inspiring assignments for students. They can build slides that are engaging with images in the platform. It is the one-stop shop. It is a place where teachers can do their work in the same way that you mentioned Harvey that supports lawyers. Magic School is kind of like the AI tool for teachers in their daily work to get things done. Is anyone using Magic School? It's just rolled out in a neighborhood of Denver. Has it gotten any traction yet?

[00:22:01] Yeah, maybe the most remarkable thing about Magic School is how well distributed it is. We have over 8 million educators who signed up for Magic School for... 8 million? Yeah, 8 million. And for scale, there are about 3.5 million educators in the United States. So that includes international folks. We have data to show that Magic School is used in just about every school and district in the nation. We go to places like the UAE and Saudi Arabia, and we find out that 60%, 70% of the teaching population in those countries are using Magic School.

[00:22:30] You can drop a pin on any developed country in the world. Just read off more countries that are some of the more active companies in Magic School. Countries. Yeah, more like just more or less all the developed world. So Europe, name a country in Europe, and Magic School probably has 50% distribution for our free product teachers, typically minimally. But we also scale with our partnerships. So the way the company makes revenue is by partnering with districts.

[00:22:58] So districts want to bring an AI tool that they can align to their own curriculum and knowledge bases, policies. You know, what schools and districts are are regulated entities. They don't want a teacher to go to a generic AI chatbot like Gemini or Claude and get, like, a generic response.

[00:23:15] They want it to be infused with their own knowledge base, the standards they intend to teach their students, whether they're in IB school, where they follow the common core standards, whether they're in Georgia and follow the Georgia state standards, or in Florida, where they have even different regulations that they have for students in schools. So Magic School provides this real enterprise solution for districts where they can infuse all their knowledge bases. They can get the kind of oversight that they need, their data privacy agreements they need, FERPA compliance. And then the second part of the platform is generative AI for students.

[00:23:45] Let me pause you there because we're going to come back to that. But I want to further establish Magic School and you. And so tell us about kind of the funding of Magic School. Where are you? Are you a pre-seed, seed? Where are you at in that process? What kind of valuation thresholds can you share with us? Because I know you're very successful. So I'm kind of teeing up a little bit of a lot of all here because I want people to know how big of a deal Magic School is already. Yeah, yeah.

[00:24:13] Magic School, we raised quite a bit of money and we've scaled a lot. We'll give a shout out to Range Ventures and the team there, which is a Colorado-based fund that led our seed round. So we raised about $2.5 million in our seed round about three years ago. We raised a Series A from Bain Capital Ventures and we raised a Series B from Valor Equity Partners.

[00:24:39] So Bain Capital, Valor Equity Partners, I mean, those are kind of top brand names. Yeah, yeah, yeah. Not bad. Antonio Gracias, who's the leader of the fund at Valor, just IPO'd SpaceX. I think they returned $100 billion to the fund or something. Ah. Incredible. So he's on our board and it's a good company to keep. And yeah, so we've scaled quite a bit. We've nearly 200 people based all around the country.

[00:25:07] And yeah, we've had both distribution, teacher love, district partnerships, but also a lot of sustainable financial scale. And, you know, that's not common in ed tech. So it's been quite a journey. So we've been able to ramp revenue and just as fast, probably the fastest ramp that any company in our space has ever had. Yeah, no doubt. And where do you do your development work?

[00:25:36] How do you get the development work done? And geographically, you're in Denver. Where is your team based? Yeah, about a third of our team is based here in the Denver metro area. And we do, we have offices that is kind of a hybrid workplace situation. So our development is done probably mostly in Denver and in the Bay. We have product managers there.

[00:26:04] We have engineers there and product management engineers in Denver. So that's where the majority of our development is done. We have some outliers in different places, but it's a remote first company. So there are folks kind of based everywhere. Our go-to-market engine, as you'd imagine, are very regionally based. So from account executives to folks who help with implementation to folks who train and support teachers and districts, they need to know the local context of that district.

[00:26:31] So as you'd imagine, Florida schools are very different from Colorado schools, are very different from New York schools or California schools. So the ability for those folks on our go-to-market team to speak the language of the district and understand the context that they're dealing with is critically important and allows our implementations to be really intentional, thoughtful, aligned to the needs of the district to help drive the priorities and challenges they're facing. And you're an endeavor entrepreneur. So you're part of the endeavor global network. Yeah.

[00:26:59] When did you make it through the endeavor process? It was relatively recently. Yeah, yeah. Probably the last six months, maybe six months ago. And we're really excited about becoming a part of the endeavor network. We're actually, I was just sharing earlier, reaching out to start utilizing the global network. We're, like I mentioned, very well distributed in the Middle East and the GCC countries.

[00:27:23] So we're looking to open our first international offices in Dubai and was just connected with the folks there to help us build our first go-to-market engine like we have regionally in the U.S., but placed in the Middle East. Yeah, I actually visited the Dubai office a couple times a few years back and visited the office in Saudi Arabia, in Turkey, Jordan, and even in Lebanon.

[00:27:50] So it's a endeavor global has got a really strong presence in the Middle East and in all the other regions that you're doing business. So you mentioned you have about 200 people. That doesn't sound like a lot of people for all of the geographies in which you are active, knowing that you need it. So it's helpful if you're going to have a concentrated presence in geography to have relationships with the school districts, with the state, with both on the sales side, but also on the operations side to help them fully leverage the product.

[00:28:21] How are you able to have your product used by so many with so few people being part of your team? Yeah, one of the superpowers of the company has been the product-led growth story and a community-based marketing effort. So, like I mentioned, Magic School, we probably got from zero to a million educators using the product in four months.

[00:28:47] And now we're sitting around eight million folks who know the product. But probably the thing about it is that it's so simple to use and simple to implement. Time to value is something that we think a lot about from the product side. And it's a consumer product in some sense. Like the teacher found it. The teacher loved it. They became champions for us at the district level. So often our go-to-market team is partnering with a localized district infrastructure that already implements technology.

[00:29:17] So we're usually running a train-to-trainer event. So an example is in Hillsborough County in Florida, one of the biggest school districts in the nation, we had a summer institute for technology coaches. And if you didn't know this, and I imagine some of your listeners might not, is in just about every – most school districts, they have like a digital technology coach who exists in every school building. And their primary responsibility is, one, keep the technology working.

[00:29:42] So like laptop breaks, whatever, overhead projector or now a smart board isn't working, get it fixed. The second is bring great technology and enable the teachers in the building to use it. So there's already this existing infrastructure we get to plug into. So we're often training those folks in large numbers to become champions and help enable Magic School to all the teachers in the building. So it's actually kind of nice. Like there's already an infrastructure in place for training.

[00:30:08] And those folks are so excited to bring Magic School to their teachers because teachers are through the moon about it. They're so grateful for it. But I've heard district folks and technology folks have told me on multiple occasions like it's the first tool I've brought to my teachers where they'll stop me in the hallway and thank me for it. Usually tech in schools is kind of like, oh, we have to do one more thing or this thing isn't working. How do I make a class in this LMS, this ancient – you know, Magic School just works.

[00:30:35] And it helps them with their day-to-day work in this really, really specific way. So it's like a joyful product. When you ask teachers about Magic School, they'll be pretty thrilled to tell you about it. Yeah. I mean, Colorado is interesting because you're headquartered here. Yeah.

[00:30:50] Colorado passed this AI law a few years back that set us apart from almost everywhere else as being a state that was putting regulatory hurdles in front of AI development in a state where AI being used for education, you know, one of the best examples in the world was right here in Colorado.

[00:31:12] And now we're – and fortunately we got that law kind of mostly corrected, which is good, but it took a period of time to do that. We're now, like, putting together political measures that are – make it hard for a company like yours to want to stay headquartered in Colorado because there are other states that would be – make it a lot better for you to be there relative to here if Colorado keeps doing what they're doing.

[00:31:40] And all that is juxtaposed over this opportunity that Colorado has. You know, I'm a big Colorado lover despite some of my frustrations. That Colorado has to say, you know what, we're going to use AI in a way that is novel and for the betterment of the entire population of Colorado.

[00:31:57] And we're going to do that for a lot of reasons, but we're equipped to do that because we have a natural partner here in Colorado called Magic School who could help guide us in our efforts to make sure that every kid in every school in Colorado is getting the benefit of a very thoughtful and proactive way of introducing AI into the education system. And I'm sure a lot of that is taking place.

[00:32:23] I know Mike Johnson, who I assume is a big advocate, has always been an advocate for education and has experience there himself, as does Michael Bennett, our senator. So maybe tell us the three or four geographies, whether they're states in the U.S. or other geographies outside the U.S.

[00:32:45] that are examples of how to fully leverage AI in education, maybe overlapping with how Magic School is used. And maybe contrast that with some maybe unnamed states or unnamed geographies who can't get out of their own way. And you look and say, oh, my gosh, they could be doing so much for their kids. How would you describe those two?

[00:33:05] Maybe call out by name the ones that are positive and maybe just let's be silent on names of those who maybe are not doing what you'd like to and probably I would like to see them do. Yeah. And, you know, we want to win them all. And we understand that folks exist on a spectrum of how they're going to adopt generative AI. But I'll hold up a couple stories from right here in Colorado that are doing extraordinary work. You should know that, like, here in Colorado we partner with nearly every system in the state, which is incredible.

[00:33:34] And you mentioned great supporters like Michael Bennett and Mike and Phil also. Phil's a great supporter. He's multiple times over. I know that Phil has a mixed record on his AI stances. But, like, multiple times over has said that, you know, Magic School is such an obvious good use case. And then mentioned us on national press as well as in more recent fundraisers. So we're grateful for Phil, too. But, I mean, here I'll give you an example from Aurora.

[00:34:03] I wonder how Phil will feel if – and I'm not saying you said you're going to do this and I'm not going to ask you because it would be unfair. But I hope this does get back to Phil. So I wonder how Phil will feel if the graduate income tax, which is on the ballot, and he is being very silent about his opinions on that. And, you know, perhaps because of the silence, the probability of passing is far higher.

[00:34:25] And if that graduate income tax passes, a lot of tech companies who are doing really well are going to – their leaders are going to feel compelled to change where their company is headquartered to somewhere other than Colorado. And with that, you know, jobs will go elsewhere, et cetera. And, you know, a company like yours and an individual like you is one that will, you know, will be most vulnerable to have to look back and say, okay, what just happened here? You know, and do I need to do something different?

[00:34:55] And if you, in going through that journey, assuming that happens, decides, you know what, I'm going to need to kind of relocate, I wonder how people like Phil and Mike Johnson, he hasn't come out – both of them are friends of mine. He hasn't come out either, and Bennett has not come out either and acknowledged the danger of us passing that graduate income tax because of the flight that will take place from tech innovators like you,

[00:35:22] who are going to look at Colorado and compare it to other states that are much more welcoming, much more – have a set of policies and a set of the way they approach successful entrepreneurs and successful businesses that will be very hard for you to say no to and stay in Colorado. How will they feel about that? I mean, you know, I'll speak to the results of this. Like what a business like Magic School starting in Colorado and staying in Colorado means for the results of school systems here.

[00:35:52] And I'll start with Aurora Public Schools. And, you know, we love being in Colorado and some of the most innovative schools and districts. You're a chair to ask for Colorado. Yeah. And you should be celebrated. You and your company and what you're doing should be celebrated. Every time we talk about tech innovation in Colorado, you know, Magic School should be near the top of that list because of what you're doing. And more than anything, it's great to build a great business. It's great to build, you know, help build the economy here. We have employees here and all that.

[00:36:18] And it's also great, you know, we can all agree that students need better outcomes in schools. They need to have better educations. And I know that everyone – It's hard. It's hard. It's hard. I mean, it's hard. We all believe that. If it was easy and there weren't tradeoffs, you know, we'd get a lot better, a lot quicker. But that's not an excuse to be bad at educating. Yeah. And Colorado, you know, can be a lot better at educating. For sure.

[00:36:43] And, you know, it's inspiring to see Colorado take the stance of being – embracing AI, right? Aurora Public Schools, one of our hardest-to-served districts in the state by demographic. It is a school district with a lot of students who qualify for free-to-introduced lunch, low-income students, folks who are English language learners, new to the country.

[00:37:06] A very important, you know, district in our state and one that – whose educational outcomes needs to improve. And huge. And serves a wildly diverse population of students. One of our absolute first partners was Aurora Public Schools. There's a story from Aurora where Aurora used one of our AI tools with their students. It's the writing feedback tool. It's this tool that students can write their essays. They write them completely on their own. No AI coach with them or anything. They just write the essay.

[00:37:36] And then they submit that essay and they get aligned to a rubric, formative feedback on their essay. So they get, like, you know, areas of strength. Like this is what you did well and how you score well on the rubric. And then here's some areas of growth. If you would do these things differently, your rubric score would improve. You would do better in your writing assignment. Then they got a chance to revise. So after they got that areas of strength and area of growth from Magic School, the AI tool that they were using, they got one opportunity to revise and then they could turn it into their teacher.

[00:38:05] They did about three practice writing sessions throughout the year in this classroom. And year-over-year growth for those students was a 28% year-over-year growth. 28% if you're in the business of schools, you know, is almost unfathomable. Like that doesn't happen in one year. I've never experienced that as a principal. It's like it's unheard of. But when you have this kind of technology. Let's go through it again. They went from where to where and what time frame?

[00:38:32] 28% year-over-year growth on their writing scores for their state standardized exams. This is widely reported. It should have gotten more press here in Colorado. But you can Google it. You can find the story. We didn't like program this. We didn't make this. 28%. 28%. And when you're in a population that is more difficult to have that kind of improvement too because of, as you said, diversity of languages, diversity of cultures is kind of a set of circumstances that makes it hard to see that type of movement. Yeah, unprecedented growth but it's unprecedented technology.

[00:39:02] There's never been in the history of our country no matter what – History of the world. Yeah, where you had this kind of technology that even if you were in the most privileged setting in the world, like the most expensive private school in the world, you never got feedback immediately after a writing assignment that was aligned to the rubric that you're eventually going to be scored on. So like, yeah, you can get absolutely astronomically different results when you get this kind of change in coaching for a student. And every kid wants to succeed. They just need to know the map of how to.

[00:39:31] And that's why I think – And they need to be met on a very individualistic set of – Yes. Because what that means for one student versus another student versus a third student could be very different depending on their personality type, their circumstances, what interests them, what is boring to them. And for a teacher to do that using tools of 10 years ago with a classroom of 25 or more kids, it's got to be really difficult. In that same Aurora classroom – I would fail miserably at that.

[00:40:00] There will be a kid who is three grade levels ahead and there will be a kid who can barely write. They need different feedback. One teacher cannot humanly give the kind of feedback that each of those kids need. Especially when they've got a larger classroom and they're distracted themselves during the school day. Yes, absolutely. So these kinds of tools can be – But with AI, that same teacher enabled and powered by the AI tools that you are building and others are building,

[00:40:28] all of a sudden they're able to give and facilitate and oversee a very customized set of educational experience for each kid. And any of those kids who wants to lean in can, I assume, find ways to lean in on terms that meet them where they are. Exactly right. And bring them forward in areas that they want to be brought forward in. That's exactly right. And in that tool, one of the most interesting things about the tool itself and the way that it's used is students get that first set of feedback back from it.

[00:40:57] And almost always that student after getting the feedback, they ask more questions. They say, hey, you said I had run-on sentences. What are some examples of run-on sentences? How can I avoid writing run-on sentences in my writing? They get really curious and they get even better because they have that tool with them. It's not taking away the teacher. It's giving the teacher superpowers. It's saying now every kid has a supervised agent that is supporting them in their writing. How incredible is that?

[00:41:24] 30 agents in a room supporting 30 kids supervised by the teacher. That's a future that is way more impactful. So let's talk about the counterargument, which I couldn't even – usually I could do a good job of framing a counterargument for someone who I don't totally agree with it. But I have trouble framing a counterargument. You mentioned New York and the new laws that they passed.

[00:41:50] Let's say there's another school district like Aurora, similar circumstance, similar size, but they had a different point of view than Aurora relative to the adoption of AI in the school. They were fearful of it. They thought it might do harm. It might make kids lazy. It might – they might not learn how to do math the old-fashioned way or how to do something else the old-fashioned way.

[00:42:15] What would be their point of view about protecting their school district from the evils of AI while they're watching school districts like Aurora kind of move forward at the pace they're moving forward? Well, I think that it's a lack of education and understanding the nuance, right?

[00:42:33] This is a problem with our national dialogue is that folks will often read the headlines and see that AI data centers are harmful to climate change and AI is going to kill us all. It might do that. That one I could understand why there might be some concerns, but in the meantime, we can't ignore it. Yeah, but in the meantime, let's use it for the really – like the powerful use cases, right? And the way we're going to really impact them. We're going to do that on AI so that AI doesn't take them over. That's right. Even speaking to that point, right?

[00:43:02] Like even if – let's say you believe that's a possibility like the folks at Anthropoc do. Say it's a 10 percent or more possibility this could happen or some really harmful outcome could happen. So if you believe that, which a lot of folks who are the same folks who architected a ban in New York City public schools, if you believe that it could also be so harmful that it might have a greater than 10 percent chance to kill us, wouldn't you want to educate the populace on this technology? Unlike this thing that could be so harmful with urgency, right?

[00:43:31] Like so some of these things – like I don't understand how you can hold both those beliefs at the same time where we're going to put our head in the sand and not let our kids know this technology exists. We're going to pretend it doesn't exist until they're in high school. Then we're going to do 45-minute AI literacy sessions. And then on the other hand, we're trying to do moratoriums at the national level on generative AI. And I love a lot of things Bernie Sanders does. This doesn't make any sense to me at all that you'd say a moratorium on AI when you – and it's because it can be so harmful.

[00:43:59] And then you're also – and one side of your mouth saying that and you're also supporting this idea that kids shouldn't learn about it when it can be potentially harmful. Shouldn't they learn about it because it is harmful? So that's the thing that we often don't live in when we talk politics. We don't live in the nuance. And even if – which I don't think there's any chance of this being the case. But even if Bernie was right and we should hold back AI, it ain't going to happen in most parts of the globe. It's not going to happen in every other state. It's not going to happen in every other country.

[00:44:26] So while you're trying to hold back whatever people you have influence over, a lot of the rest of the world is going to do the opposite. And who do you think is going to win that race? Yeah. Not everyone's got held back. Yeah. And I think that the thing is what you choose to do in your local community, I'm going to support as long as it includes educating your kids about generative AI. It could be that we don't see a great use case right now for generative AI for kids in K through 5. That's a – there's a reasonable reason to say that we're not going to use it very much.

[00:44:55] But we're going to let them know that they might experience deepfakes in the world and they should know that not everything is real. They should understand this technology exists and there's good reasons to use it, bad reasons to use it. There's a way – even those early grade levels to help kids, even if you believe it can be harmful, you still need to teach about it. You can't put your head in the sand and say it doesn't exist. And there is a risk of cognitive offloading for kids if it's used in improper ways. If you're saying write the essay for me, I'm not for that. I think kids need to preserve –

[00:45:50] they're not going to even realize how primitive that output is and that not only can anyone do that, but most everyone else has already done that multiple times in the past and they learned that that's not a finished product. That's barely even getting started. And boy, if you spend your first five years barely using AI, you're not going to know what you don't know. You're going to be so ill-prepared. Absolutely. I'll cite the story.

[00:46:18] A friend of yours and mine, Brian Leach's daughter. I'll save her name from this conversation. She to me is what I think we all want for our kids when they're experiencing generative AI in the world. I went to a Nuggets game with Brian since he's a big Nuggets fan and a sponsor of the Jersey. And his daughter was there. And we start talking about AI. She finds out what I do. She knows I'm a principal. She finds out that I'm running this company, Magic School.

[00:46:48] And she says, oh, we're using Magic School in school. And I was like, oh, amazing. Like tell me more about how you're using Magic School. And she describes the writing feedback tool I told you about that her history teacher was using the writing feedback tool with their class. And she's – at the time, she was in 10th grade. And she could very clearly articulate the value of it. She was like, yeah, it's so nice to be able to revise my paper with the feedback before I have to turn it in for a grade. It's so helpful and personalized. Like it's so useful. And I ask her a bit more because I'm just curious.

[00:47:17] You know, what is a 10th grader thinking about AI? And you have a dad who's in tech. So he's probably talked to you a bit about it. And she tells me – well, I was like, what's really happening? Like what are your friends doing with it? And she's like, well, I know a lot of them are using it like as a shortcut. Some of them are using it just, you know, when they get home. They're having their essays written with it or, you know, that happens. But I was like, well, what are you doing? And she's like, well, I kind of use it sparingly. I think about the reasons it can be really helpful to me.

[00:47:45] And I love when my teacher uses, you know, the magic school tool with us because it's so clearly useful. But I do know that like I'm going to be tested on a lot of the work that I do without AI. So I need to preserve my thinking. So I try to use it really thoughtfully. But right now I've been spending – I haven't spent a ton of time with it. But I'm excited about the good use cases. And it was a nuanced answer. It was really thoughtful. People don't believe me when I tell that story.

[00:48:14] People don't believe that someone could be so nuanced and thoughtful. But I was like, well, like not only does she have a father who obviously is like working in tech and understands that this is a part of the future. She also has a school that is progressive enough to embrace AI for positive use cases. So she has this like ecosystem around her who's – they're telling her both like, hey, you don't have to use it for everything. Don't use it for that cognitive offload part. You're going to have to take these tests on your own. We want to preserve your thinking. Your voice matters. But also there's some really good use cases.

[00:48:44] Don't we want all of our kids to have the kind of agency that Brian's daughter has? Don't we want all of our kids to be able to say, hey, this is when to use it, when not to use it and get that experience when they're in high school? That's unfortunately the exception of the rule right now, right? Brian's daughter has this leg up because she's gotten all this education. Every kid deserves that, right? That's what equity looks like. Yeah. And the answers to these questions are hard and will be discovered over time.

[00:49:13] A couple months ago, I had to build a spreadsheet and an important one because we were making a decision that had to do with the $60 million purchase of a building and kind of what to do with it and whether it was good price, not a good price, how that would look over time. Well, when I built that spreadsheet, I didn't use a spreadsheet. I just used Claude.

[00:49:35] And I worked with Claude to flush it out and build it in a way that you could see and understand all the assumptions that went into it. But you could also be researching each assumption while you're building it. Go dive deep into this. Give me all the examples. What's the right number? Why? Give me your logic. And in a short period of time, I built, again, a very important use case. This wasn't just screwing around.

[00:50:02] This was to make a big decision and want to be as thorough as possible. So with a very small amount of time, built a very thorough multi-year financial analysis and was able to move very quickly with the project and one that I was able to have really high confidence in the result without ever seeing or touching anything that looked like a spreadsheet. You know, if I would have done that even five years earlier, one, I wouldn't have done it because I wouldn't have spent the time to do it. But two, we wouldn't have gotten the deal done.

[00:50:31] I wouldn't have had the confidence and conviction of what we were doing. And I wouldn't have had it documented in a way that it could be used by others. And they said, okay, let me see the model. Let me see all the assumptions and logically on all the assumptions. And that could be given over and they could probe it over and over again and we could update it on a flash. So, you know, was the fact that, you know, I knew how to use a spreadsheet in a more traditional way in the past, was that helpful? Sure.

[00:50:57] But would it have been worth it for me if I hadn't already had years of experience using spreadsheets? I have to go back and say, well, I better learn how to use, you know, Lotus 1-2-3 because that was part of my experience that led to Microsoft Excel. And I need to recreate all that and spend a bunch of years doing so before I could get the benefit of Quad building on a model. I don't think so. I think that would be an enormous space of time if I was, you know, earlier in my career and starting now.

[00:51:23] I still need to understand and have intuition about kind of, you know, how to do evaluation analysis, how to do diligence, how to probe and make sure the assumptions, how to know what I'm looking for. But I probably would have learned that in an AI context if I was young and learning it now, not had to go backwards in time and say, well, I better do that with a pencil and a piece of paper and a calculator first so I could build up my skills on how to use quads. So, I don't know.

[00:51:52] I guess I would be in the far end of the extreme of advocacy of, you know, even an example with Brian Leach's daughter is don't overthink this. Yeah. Use AI every opportunity you have. Still be responsible for the outcome. Be responsible for understanding what you're doing. But do that in a way that fully harnesses the power of AI in every sense of the word as opposed to saying, man, I'm going to use it more selectively. Yeah.

[00:52:17] What you described, I think, is what I think the future of a college assignment could be, right, is investigate this with AI. Here's a financial model. And you're going to upload the financial model and you're going to have to investigate whether it's credible or not. And you're going to do the things that you did in that. You're going to say, what were your assumptions you made when you made this calculation? I've got friends who talk about their 10 or 11 or 12-year-olds coding with quad or the equivalent. So I don't know.

[00:52:45] I don't know if it's true because, you know, your parents for, you know, hundreds of generations want to talk about how great their kid is. But there's certainly some elements, the truth in that. And my question would be, man, if your kid's already learning how to do things that you'd call coding at the age of 10, 11, or 12, you know, build a game, you know, help with their homework assignment, whatever they're using it for.

[00:53:12] Or, you know, I would think they're going to have legs up on others very, very quickly and very profoundly. Yeah. That reminds me of a story from a district-sponsored event I was at recently. The district, my name is actually a partner of ours. And they're wonderful and progressive. But I think you should know this is probably – this is in a community that you'd capture as probably – sees the headlines around, like, doomerism with AI. They had a really incredible event.

[00:53:39] It was an AI event for kids across the district who were interested in learning more about AI. And it was on one of the days off that students had traditionally. So every kid wasn't in school that day. So these kids all came to the central office for the district to get this AI literacy day. It was really cool, well done by the district, really thoughtful. In one of the activities, and I got to be a part of it, I spoke for a bit. And then there was activities for the kids to, like, you know, do different things with AI and utilize their knowledge.

[00:54:07] One of the activities was solve a problem in your community with AI. And they had this kind of, like, sheet where they had to, like, think about the problem and, you know, go through their thinking. And mind you, this is a self-selected group of kids. They signed up for this day. They were really excited about AI. And they came to this session. There was a kid who was kind of sitting by himself and didn't have a partner. So I went over to that kid and I sat next to him. And I said, you know, tell me about your idea. And probably age range.

[00:54:36] This kid was probably a freshman in high school, 13 years old or so. And he shows me his paper. And it says, I'm trying to stop AI. I actually want it to stop because I want to be an engineer. I want to code things. I want to, like, build video games and stuff. And I hear that AI is going to take my job. So that's his project. It's not the assignment, but that's his project. And that's how limited information kids have as an illustration. I start telling him about, like, I'm like, oh, no, no, no.

[00:55:06] It's not going to take your job at all. More engineering jobs are available today than they were before AI. Like, jobs are growing in this space. This is really exciting. And also I start telling him about things like Replay and Lovable and, like, you know, Cloud Code and all the things you can do to build your own software. And he turns from, like, a person who's, like, trying to stop this and thinks it's going to take his job to, wait, what, I can make my own video game now? Like, yeah, you can do that today. Like, go home. These apps are free. Try it out. And he gets so excited.

[00:55:32] And, again, like, that's a kid who's self-selected into an AI day who wants to be an engineer. That's, like, the kind of misinformation that he's getting. Yeah, let's talk about the misinformation because that's concerning to me as well. When I hear, and my exposure is more to kind of college kids than younger, where whatever set of influences are going into them, they're having all these negative thoughts about AI and how they're going to resist AI.

[00:56:01] And it's like a chip on their shoulder with AI. And they're not going to, like, learn AI. And AI is, you know, the root of all evil and this and that. And I'm like, you better get over this really quick if you don't want to be left behind. And whatever influences there are that are telling you this stuff, it's going to take away all the jobs. It's going to do this and it's going to do that. And I'm going to show AI. I'm going to show it by not doing AI. It's like, really? You know, why don't you stop using the Internet as well and see how well that works for you?

[00:56:27] But if they lose those, you know, junior year and senior year of college, you know, spending their time coming up with reasons why AI is bad and they're going to resist it, you know, they're going to be, before they know it, they're going to be 25 or 26 years old and be looking around saying, why am I so far behind in my career and why is my trajectory, like, flat or going the wrong direction? It's not because that's the way the world is. You're going to see a bunch of your peer group who are way, way ahead of you.

[00:56:56] You're going to see kids who are two years younger than you, three years younger than you, who are so much more advanced in their capabilities to contribute in the real world. So, and, you know, and you're going to feel really good about what, that you are a rebel without a cause against AI? It's like, even if, and this is from someone who places legitimacy in that there are real world concerns that we should have relative to AI,

[00:57:22] including, and maybe at the top of the list, this AI divide that even if we're doing all the things right with AI, which we're not, that AI divide is still going to be really profound between those who have more access to, you know, to things because of the circumstances they're growing up in to people who grew up in less advantaged situations. Even if we were doing everything right, that's going to be hard to overcome. But on top of that, we have this added and extreme problem of, you know,

[00:57:52] certain school districts suppressing AI intentionally. I mean, the price that a generation or more of, you know, of those affected by it are going to have to pay is going to be enormous. Yeah. I couldn't agree more. I think there's real, real complexities, real things that we need to, to consider real, real concerns. I think that a lot of the, the bans and the limitations are, they come from like a place of, a good place. They come from a place of, of caution and safety.

[00:58:20] And again, the information that you're fed, but, but I also, I've always been, I mean, my whole life, I've been like a tech technologist. I've been like a techno optimist. I would be, if I wasn't doing this, even as a teacher, I would be always trying whatever new tech was out there. And I'm telling, when I tell folks, I'm like, you know, I can be a trustworthy person to a lot of folks who are skeptical. Cause I'm like, I worked in public schools. I worked in, in these things that, that usually don't associate with someone who works in tech.

[00:58:48] I'm like, no, no, this is not, this, this is a real unlock for folks who can really, really change the trajectory of a kid. I mean, talk about the equity piece. If you're in New York city public schools, you don't get access to an AI tool or you're not taught how to use one at all. Um, and you're an immigrant family like mine was, uh, who had no idea how to navigate the American public college system or, or the FAFSA or all the things you need to do to apply and get into college.

[00:59:14] Um, I was relying on two counselors for a high school graduating class of about 1700 kids. Uh, or I could have gone to an AI chat bot and said, how do I fill out the FAFSA? This is my parents' circumstances. What colleges would be good for someone who wants to major in this? Uh, and just getting those answers right there. Um, and then accelerating my own ability to get opportunity and, um, pursue the American dream like so many desire. Right. Um, every kid deserves that access.

[00:59:42] Every kid, kid, kid needs to, to be able to have those skills and democratize the access to information. Um, uh, there are these such obvious use cases. It breaks my heart that like, you know, a kid in that kid was told, was said, the only thing I can do is ban AI. The only thing I can do is stop this AI thing from happening. Uh, when the actual answer for a, somebody wants to be a software engineer is that you can actually build a game today instead.

[01:00:06] You can actually do things that you, you thought you'd have to wait until you're in college to do, uh, today with these tools. And you can be a creator. You can, can pursue things in your imagination that you thought you'd have to wait years for now. Um, that kid went from scared to inspired. Like, uh, I think that, that we can get to a place if we, if we start democratizing this, this technology and helping people use it for the right reasons. Um, but they're not going to use it for the right reasons or if they're not taught how to. Um, yeah.

[01:00:37] And they're not going to use it for the right reasons if they also don't use it for the wrong reasons. You got to learn. Plus to kind of come together. Maybe, you know, the proportion of right versus wrong is where you have to work. So let's talk about, I can't remember the name of the, the, the new school concept that I think is coming out of Austin. Can you tell us a little bit about that and how that fits into the overall scheme?

[01:00:59] I mean, that, you know, that's an example of those who have access in part because they have access to financial resources, but also they're in geographies where it's, you know, being rolled out. You know, if that proves to be, you know, a model that is very advantageous, it will grow the gap even more. And there's certainly proponents of that. And there's certainly those who are going to resist that at all costs. Uh, so what am I talking about?

[01:01:28] What is, uh, I know enough to know just enough to be dangerous, but. Yeah. Yeah. So there, there's a new school model that's kind of catching fire around the country. And I will not purport to be an expert on alpha school. Uh, this network of, of private schools around the country. Yeah, it's called alpha school. Um, but I'll, I'll tell you what I do know. And I'll tell you where I think it, it, uh, I'll offer both like a critique of it and I'll offer why I think a lot of families are really interested in it.

[01:01:55] And, and the tuition at the school can, can be up to $50,000 a year for a kid. Um, so not public schools. Um, but, but the, the general premise is kids spend two or three hours a day learning with, with AI tools. Um, that are, I think mysterious at this point of what set of tools they use to make this work. Um, and, uh, with a kid with an incredible agency, uh, that might be a really great learning experience where they can ask the next question, continue to grow. And I think they use a series of tools.

[01:02:22] I don't really know how that works with them, but it's, you know, they, they purport really good results in, in some of their schools. Uh, and then the second half of the day is experiential. So it's, you know, we're going to go into the real world and we're going to learn, uh, in a project-based format, um, the physics of a skate park and like how kids are doing, um, uh, you know, the, um, the tricks that they do at a skate park. Or, hey, we're going to use, maybe, maybe in the second part of the day, it's not AI based at all, but sometimes it is AI based.

[01:02:49] Say we're going to use, uh, Replit to build, you know, build software. Um, so, uh, I'll tell you from the perspective of well-to-do families, if I'm around like folks who hang around UDAN, like LPs and stuff like that, like they will, and they have kids, they'll always ask me what I think about Alpha School. Um, because it's, they're considering it as an option. I had a really interesting conversation recently here in Boulder, uh, of a parent who was considering it as an option because they're, they're thinking about opening a school here, I believe. Uh, and the thing that she told me, I was like, why are you considering it?

[01:03:19] Because you have, you have every opportunity. Like you could send your kid anywhere. Like, you know, you, you, you have means you could send them to the most prestigious private school here. Um, Boulder Valley is great schools for, for the district schools. Um, what, what made you interested in Alpha School? I was just genuinely curious. Um, and she was like, well, I'm using AI and it's, I'm learning so much. I'm using it as a learning tool. I like you just described using this as a learning tool when you talked about the model that you created.

[01:03:45] And I just know from my own formative experience that it can help you learn really fast. I don't have to buy into the idea that AI can, can be. I'll, I'll, you know, put an explanation point on that. I mean, the, the, the opportunity to learn and understand and go deep and probe in areas that you are interested in, even if they're quirky areas, you know. And if you really want to know, you know, and learn and discover, use the word imagination a few times.

[01:04:12] I think imagination is maybe the most important word right now of any word when it comes to AI. The ability to take what you might imagine and do something with it that you want to do with it. And I mean, and maybe that's where my passion comes from. I haven't heard this story that you're sharing, but maybe that's where my passion comes from. Like, I can't believe certain people are going to not get the benefit of this.

[01:04:36] Not because of, of things that are hard to overcome, which, you know, some of the socioeconomic are hurdles that are hard to overcome. But then on top of that, there's an intent to not get exposed to that because of some kind of fear factor or some other type of discretionary decision making that makes no sense to me. Yeah. And this mother works in tech, so she's deep in, right? So she gets it. She gets it. And she's like, well, I mean, I'm using, I'm using like a, I'm using a recorder to record my meetings that I'm having.

[01:05:04] And I'm asking for feedback from those recordings. And I'm learning, like, I'm growing. It's like telling me that I could be better in my questioning. It's telling me all these different things. And she's so inspired by, like, how much she's learning with it. So she's like, well, that's one reason is, like, I don't have to guess if AI can help people learn. I've lived helping me learn. And I'm improving at my own job with this. So that's one reason it's interesting. And then she also talked about this experiential side of the school. And I was like, that is interesting, right?

[01:05:32] It's like a very unique perspective that you bring to this. So I'll offer the critique. And that's that. Let me add one more thing that I put my, you know, you were telling the story and I was putting myself in her situation. You know, if I had, you know, had my own kids of that age. The other thing that strikes me is I'm partnering with an entity that is committed to figure out how to leverage AI for the betterment of that next generation.

[01:06:01] So whether the exact thing they're doing right now is the exact right answer, I'm trusting this entity is doing everything it can conceive of to accelerate the AI experience for the people who are part of their program. That would mean a lot to me as well versus putting my kid in an environment where the level of conviction about commitment toward AI is more mixed. Like, well, we're kind of like this. We're not so sure about that. And this and that.

[01:06:30] It's like I'm going to bet with people who are trying to go all in with AI. Yeah. And that's to talk back to the school board member story. She was like, it wasn't a good enough answer for my kids for the district to not have a strategy for AI when I'm using it in my workplace and similar to the conversation I had here in Boulder. But the critique I'll offer of that is the idea of sitting a kid in front of a computer. One, it probably is like a really high agency kid who's going to be able to really utilize this.

[01:06:59] You know, they say they believe in relationships between the adults in the building who I think they call coaches rather than teachers. Coaches are another word I'll remember at some point. And they're kind of like they incentivize the kids. They kind of build relationships with them. They say, hey, we understand relationships are important in schools. And I definitely know that's important from my experience.

[01:07:20] Now, my critique of the model is that like the computer independently teaching a student in K-12, even with a coach, is I think it misses the idea that one of the reasons you relate to a mentor or a teacher, which are typically, you know, you could use the words interchangeably sometimes, is because they can impart some kind of wisdom upon you. And they know their content. They know their subject. They're someone that you trust. Some of that is the magic of schools.

[01:07:49] You know, it's really hard to depart from just say that you're completely – you can't offer any content as a teacher and completely separate the AI. So the way that we've built our student tool is saying that, yeah, like a teacher supervising these agents, the teacher is still in control, making sure that the agents are orchestrated in a way that the learning is happening. And they're still seen as like the head teacher, right, of these agents that are supporting the classroom.

[01:08:13] So I think that just putting it in front of the kids, I think that does work and it's worked historically for really high agency students. But that's like the one person of the – Yeah, it's obvious to me that the role of someone – the role that they would describe using the word coach and what you just described using the word teacher or mentor and one who's controlling agents, that those are two – those don't strike me as two necessarily different things. Yeah.

[01:08:40] Different set of words, but I would imagine – I don't know, but I imagine what they call a coach, you know, is filling a responsibility to the kids that are – that they're responsible for. Yeah. And they, you know, are viewing kind of AI as a tool set that they're using and are overseeing and exposing that to the kids.

[01:09:02] And there's – and I'm not sure that that would be that different than what you're envisioning the role of a teacher who's using magic school that you also use the word mentor for, that they're using this as kind of a set of agents that is able to kind of meet the kids in a more customized way under the supervision, oversight, and leadership of the teacher. And, you know, without getting the hyper-specific special model, mind you, again, I know we're both not experts on Alpha School.

[01:09:31] I've listened to like a podcast about Alpha School, so that's about the limitations of my knowledge. But do I think that there is some version of a school that is much more AI progressive, which is what Alpha School intends to be – take away what the specific model is, that is much more progressive about using AI to do what that parent wants for their kid who's going to Alpha School?

[01:09:52] Is that a possibility in the future? Is that something that is likely a better scenario than the one that we have now where there's one teacher, 30 kids, same materials, radically different levels, radically different interests? Almost certainly. There's definitely a better way to do schooling with this technology. And just like a great coder will be so much more empowered because of AI, so they're not going to be replaced. They're going to be amplified in terms of capability.

[01:10:21] I would think the same is true with a great teacher. A great teacher enabled by AI, the impact they could have on students has got to be 10x the impact they could have had pre-AI. Yeah, and I have a story about that. That's actually – that is absolutely true, and it's evidence-based. So – Oh, I'm sorry to interrupt, but let me also make clear as I say that what I'm not saying is that same teacher could teach 10 times more students. Yeah, yeah, yeah, yeah.

[01:10:48] I'm saying the impact of that teacher could be 10x on the students. Yes. And now maybe the best teachers impact more students. Yes. Because of that. But the translation of that is the education gets far better. The student experience gets far better. They get much more prepared for life.

[01:11:07] Their prospects of living a successful life hopefully get much higher because this great teacher was empowered to contribute toward the life experience to them in ways that would have been very difficult without having access to AI tools. Yeah, we're on the same page. Teachers don't need any more responsibilities, but can we make them more impactful on the group of students that they serve? Absolutely. And so Aurora Public Schools, I'll go back to that example.

[01:11:34] Another really exciting story out of Aurora Public Schools is – the way the story goes is the director of technology in the district, director of innovative technology, calls up the regional superintendent after they get their state results back for the last year. There's a metric in the state that folks who aren't super acquainted with education won't know and the folks who are leaders in education will know. It's called MGP, median growth percentile.

[01:11:59] It's the – it's actually a teacher-associated metric that identifies how much a teacher has grown that student, an individual student, year over year in their classroom. So like based on their last test results versus these test results and their expected growth, did you outperform their expected growth of this child in this classroom? So like a 50 would be like – actually, I'm not going to – I think it might be a 30, but we'll take the numbers out.

[01:12:23] But the number itself is like if they get the average growth, it would be like, OK, you did a good job, but you didn't do an exceptional job. They didn't grow more than they were expected to grow. And if they grew beneath that number, it would be like, OK, that kid fell behind, right? They didn't even meet their expected growth. And it's a really rigorous number. It's based on the individual child. It's not just like – it's like, hey, how much did you think this child would grow?

[01:12:48] So in Aurora, there's a direct correlation between the teachers who are using Magic School and the teachers who had the highest average MGP scores. They compared lists of their power users in Magic School, the folks who are using it the most, the same way that you might as a coder or a technology officer in a company see which of my engineers are using AI the most. And are they producing more PRs? Are they driving more value to the business? They did the same thing.

[01:13:15] They said, hey, which of the teachers are using AI the most? And are their results better? Are kids growing more in their classrooms? And the answer was like a one-to-one correlation. Like the story they told – and Stanford is studying them right now because of this – is that the technology director took the list of the top users of Magic School. We give them like usage reporting. And he started reading the list.

[01:13:35] And then the academic leader in the district, the regional superintendent, had a list of the absolute – like ranked, the top teachers in the district by MGP score who had the highest growth in their classrooms. And it was one-to-one. It was like, all right, this is the teacher who had the best MGP in the district. And then he said, OK, that's in our top 20. Is that – this is the number two? Yep, that's number three. Like it was just back and forth. They found this out. Stanford caught winded it.

[01:14:00] And they're studying those teachers to not just find out, hey, you used AI a lot, but how did you use it so we can spread this to the rest of the country? And I imagine that if you went back five years ago and you were doing the same thing and you discovered that the teachers have performed the best, so this is pre-AI, that the difference that they made in their students was fairly modestly better than teachers who weren't in that top edge line.

[01:14:30] Like there wasn't – there was a disparity, but not – but if you compared that disparity to post-AI, the disparity probably grew dramatically. I don't know that I haven't seen the data, but I would be shocked if it was like that would not have been possible for certain teachers to outperform other teachers relative to how the kids advanced.

[01:14:55] That wouldn't have been – those stats didn't exist of that magnitude five years ago, but amplified by AI, you're probably seeing huge jumps. Now, that means that all the boats can rise. Yes, absolutely. Even the teachers who are in the middle of the pack are probably perform – have the opportunity to perform much better for the students than the middle pack teachers would have five years ago before AI. That's right. I would love to see if those stats play out.

[01:15:22] And, you know, not only are those kids getting a better education, growing more, right, I bet that teacher is also getting home at an earlier hour, walking their dog, wanting to stay in the profession over the long term. Well, let's talk about that because I brought this up early in the conversation as part of – I imagine for a lot of teachers, this is a very stressful time because, you know, a lot of teachers aren't tech forward.

[01:15:49] Or you were, but, you know, I could rat off a bunch of people I know who are teachers. And they aren't the most tech forward people I know. In fact, they're – you know, chose teaching in part because, you know, they weren't tech forward. And they're not the ones who like stability. They like doing the same thing next year that they did this year and this year like they did last year without too much change, without too much – you know, that's just why they chose the teaching profession.

[01:16:15] And now all of a sudden they're faced with this enormous opportunity of leveraging AI. But at the same time, it's like everything has to change. What they do has to change. How to use AI. The fact that they got to get comfortable with even the language of AI, let alone all the things that you and I have talked about. But I imagine that is, you know, that is a big ask of them and one that is causing them not to go home a little earlier and walk the dog and get a good night's sleep.

[01:16:44] But huge amounts of anxiety, huge amounts of maybe even guilt that they're worried that, you know, their inability to adapt to AI like others are might be affecting the kids in their classroom. I got to think that's a real world transition that they're going to have to go through. Yeah. You know, you should know that this is happening quicker than you might imagine.

[01:17:08] You know, Magic School as a single entity company partners with schools and districts that represent one in five children in America. That means that they are paying us and they have elected to partner with a company like Magic School to enable their teachers and eventually their students with generative AI. So that's some optimism around this. And like I told you, teachers are thrilled about Magic School. Like this is so helpful. So the way I think about it, I thought about this from the very start of the company, was that if we all see the opportunity with generative AI for kids, it's so exciting.

[01:17:38] But if we're going to use AI with kids, we need to have AI-enabled adults in the building, teachers who know how this technology works. They've used it themselves. They feel literate with it. So Magic School gives them this really easy entry point. And it starts with simple things like I'm a history teacher and I'm teaching civics. And there's this really great current event that I want to teach to my kids, but I don't have any curriculum prepared for it. I know it aligns to this. I don't know how I'm going to do this. And tomorrow I have to teach. Like I don't have any time to plan this.

[01:18:08] I'm just going to use the lesson I used last year, even though I know this one that's happening in current events would be so much more relevant to my kids. It'd be so much more inspired by it. But I'm just going to teach what I taught last year because it's easier. What if Magic School made it easier? Yeah, easier sounds like lazy, but you're not saying that. What you're saying is to adapt to the new and lazy generation is a bulk of work. And there's not – you're in class all day. You're teaching the kids. You have to score their tests and you have to get –

[01:18:38] So the ability to like introduce a whole new topic is, you know, pre-AI is – there's not times in the day that you've got your own family to take care of and they want to do that. And I guess that's kind of what I mean by guilt. You know you have this opportunity to convey in pre-AI. I mean how would you do that? Yeah, yeah. Teachers are heroes. By no means is it – like do they take the easy way out? That's not – I'm not sure. It's clear of that. But it is a function of time, right? It's like you're – there's this persistent level of anxiety. When I was a teacher, right, it's like, oh, like,

[01:19:08] I made this really great lesson but that kid in my class, I still haven't been able to reach. They're still too far behind grade level. I still haven't been able to reach them and I don't know how to because I don't have the materials to support them or – I could have made a more inspiring in lesson today but I had to stay after for the sport I coach and I had an IEP meeting and I had all these different things. And I knew that I could have made a killer lesson if I had the time today, right? But now I can. Now I can differentiate this assignment for a student who's a couple grade levels behind.

[01:19:35] I can create an extension activity for the kid who's ahead who's bored in class. I can now do that because Generative A can help me do that. I can create that really inspiring, engaging lesson that's connected to today's events I know my kid is going to be excited about in the classroom versus the sterile one from the textbook. I can do those things now. So all those anxieties that have persisted now become this really joyful thing. They can lean into the intentions they've always had.

[01:20:02] I appreciate how positive you are about this and there's obviously a lot of reasons to be positive. But it's also got to be because for every example you gave, the problem with AI and we experience this and I experience this in my daily life now is there's 10 more things you can do and want to do.

[01:20:21] But you can't get to them all and you know how impactful those could be because you know the capability, the power that's in your being that you could, if you had just some more time, you could do this. And yeah, I could customize for those three kids because they need it, but those next four need it too.

[01:20:40] But I don't have time for all seven and there's got to be like what you guys are doing, what Magic School is doing is helping them through that process, giving them a set of tools that ease them into it and amplify their ability to have those impacts. But at the same time, you know, this is just the reality of the situation.

[01:21:01] And I think it's important to acknowledge that, you know, the teachers are going to – a lot is going to be asked of them during this transition. It's going to be for the betterment of the education of our kids, but it's going to be daunting. Yeah. I'll get on a soapbox here. I think the – I couldn't agree with you more. Teachers deserve training. They need enablement. They need support. They need time to be able to learn these tools and learn how to use them well if we want that for our kids, which I think we do. We do.

[01:21:31] Definitely do. Yeah. And, you know, when we talk about folks with privilege, right, and I've worked in schools with – I grew up as a low-income kid. I worked in schools that were primarily serving low-income families. And I also happen to – I'm a venture-backed founder. I meet a lot of folks with a lot of privilege now in certain circles. And like I told you, we have a family I talked about who's thinking about alpha school as infinite money, could spend any amount on their child's education.

[01:21:57] When you think about like American values and you think about budget lines for families and people who have means, my – I have a deep belief that if you asked a parent for their own children, if you had means, what are your infinite budget lines for your kids? Number one, the infinite budget line you'd spend any amount of money on is healthcare. Number two is education.

[01:22:21] So it is not surprising that in New York City, the private schools in Brooklyn and New York are now like $90,000, $95,000 a kid for tuition. You could raise that to $150,000. For folks with privilege, it's their kid's education. They would spend any amount of money on it. We don't invest enough in public education. I mean fundamentally like – it's our personal values. OK. Well, let's not –

[01:22:50] But it is our personal – Throw more money yet. It's not going to make a – But our personal values – But it's like – On a personal level – and I'm going to argue with you on this. On a personal level, that's true. The personal values would be – Americans' personal values would be if I had the means, I would spend any amount of money on education for my child. If you thought it was going to good use. Of course. Of course. Well, you say of course, but that's the problem with saying we don't spend enough on public education. It's like spending more going to make it better? It hasn't worked so far. Not always.

[01:23:20] But if we thought spending more would make it better, I think there would be huge support. The problem is for people like me who feel horrible about the job we do as society in educating our students, particularly those who don't have access to things like expensive private education, you want it to be so much better. But we also see – I also see, at least I think I see – that it's not for lack of spending a lot of money.

[01:23:47] We keep throwing more money at it, but what we see the outcome being is a bunch of bureaucracy and kind of constraints and restraints, including some of the examples about, well, we're not going to teach our kids AI. It's like, you want us to what, increase your budget when you don't even want to teach them AI? I mean, that seems crazy.

[01:24:06] So I think there needs to be a two-way street there and say we need to make our education a lot better and we need to realize part of that may be needing access to more resource. But part of it needs to be things that don't take more money, they take reform, true reform, true leaning in. And some combination there is the answer. To be clear, I don't think the cavalry is coming for a ton of money in public education. I will stand on my soapbox and believe that there could be well-used money in education that's not there.

[01:24:34] Oh, I agree with that. I'm not fussing with that at all. But I do think that like, you know, when you are a family who sent your kid to a private school that is, you know, whatever, charging, whatever, you can go to the school and you can tell them, I need my child to have this kind of education. Those schools are not slowing down. The gap between the kids who go to the private schools and who are not slowing down their AI acceleration with students and the gap between those and the kids in the public schools will only grow.

[01:25:01] If we don't do something about the kids in the public schools and if it's funding to enable teachers, if it's something like that, that's fantastic. But my point is, is that like if we value our kids' education as much as we do on a personal level, we need to start voting like we do. We need to start, you know, demanding the kinds of things that we know are going to accelerate our kids at the school board or those kinds of places. And it's not always money is a tool.

[01:25:31] It's not the end-all solution. But I do think that we as a society need to think really, really hard about the way that we serve kids because it's very clear on a personal level that people will do anything. I like why you said that, the way we serve kids because, yeah, I think we're not, we're coming way short of what we should be doing for a whole bunch of complicated reasons. The other way when we think about like these alpha skills and how much money is being spent per student by those with means,

[01:26:01] there's also an effect of that where they're carrying the load. You can look at it that they're carrying a bigger load in the fixed cost investment of how to learn how to use AI to improve education. That now there's a little bit of two way like, yeah, but their kids get the benefit of that. Yeah, but, you know, everyone's going to benefit from that over time.

[01:26:29] You know, everyone's going to benefit from, you know, from the discovery and the building of tools, the building of software, AI capabilities, approaches that deliver a better education. Because people are going to spend that kind of money are going to demand better outcomes. Yes. But what gets built as part of that is then going to proliferate through not just, you know, all communities here in the country we live in, but countries throughout the world.

[01:26:55] And, you know, I think we, you know, and there's, it's a tricky topic because, you know, what's fair and what isn't fair. But I think there needs to be more of a partnership mindset around that and more of a, you know, we need to let them spend a bunch of money, waste a bunch of it on bad ideas so that they can discover what the really good ideas are. But the rest of us need to pay really close attention to what's working there. And instead of, like, you know, throwing stones at it and say, well, that's, you know, because they have money and they can waste it. Like, what are they doing that's right?

[01:27:24] And how do we be fast followers and rolling that through the rest of our school systems and take advantage of the fact that they're spending a lot on education for their kids to discover how to leverage AI and make sure the rest of our kids get the benefit of that as quickly as possible. So, which something like magic school is a great tool to help make that happen. Yeah. I couldn't agree with you more.

[01:27:48] I think that people, people want me to, like, when they have asked me the question around alpha school, they want me to have, like, a real anti or a real pro or some, like, really spicy take. And my take is the one you have, which is I love that alpha school exists. I love that some kids are, you know, they're being really on the edge of being progressive so that, you know, the rest of the system can learn from the good takeaways from it.

[01:28:09] But I think that there's a, you know, you're kind of in the middle of that because I'm sure you face a lot of pressures by those who you want to serve. You want them to be your customers and you've got to walk a fine line.

[01:28:23] But one step removed from you is a board member of a school district or a superintendent of a school system and the pressure they face just to explain why what alpha schools or whatever they're called is what they're doing is inappropriate and what we're doing is better. That political or community pressure they face has got to be enormous. Yes.

[01:28:45] They have to almost ignore the existence of those or when it's brought up, you know, separate themselves from it as opposed to saying, yeah, I got two people on my staff and all they do is study everything those alpha schools and schools like them are doing to try to discover what's really working so that we can roll out on our school system. And they would probably be, you know, they'd probably be like, you know, beaten with a stick if they came back and told them they were spending money on people to learn from the alpha schools so we can roll it into our school district. Yeah. Like, what are you doing? Yeah.

[01:29:14] When we politicize these things, we don't come up with nuanced answers. It's like, no, like we should be learning from those kinds of systems, you know, and we should also decide what we want and don't want from them. And, you know, that should be an open dialogue. But, you know, that answer I think – I actually think most families in a district, if they found out that, yeah, we have somebody in the district dedicated to learning about innovative school models, alpha school or not, because we want to learn from what's working with kids and we think that the system needs to improve.

[01:29:41] I think that a parent would stand up and cheer for that school district leader and say, yeah, we do know it's not working. The results in our nation say it's not working. We have like some of the worst test results in this past year for the NAEP, which is national – the national exam for kids that kind of benchmarks where we are versus prior years. We had one of our worst reported NAEP scores based on grade levels in this past year.

[01:30:08] Like nobody – it's – no matter what side you're on, these results are not OK. So if you're telling me we're going to do the same thing and somehow that's going to result in different outcomes, that's not the answer. Everyone knows innovation is needed. So, yeah, we should have – we should be celebrating the folks who are trying to push the ball forward, learn from them, and recognize not everything that they do is for everyone, even the kids who go there.

[01:30:33] I often tell parents when they ask me for advice on that, I say you should know that you're electing your child into an experiment, right? Like it's not as safe as sending your kid to the more traditional school. Like it's an experiment. There can be a lot of things that go wrong with a new school. I founded a new school. I know that intimately. If you're willing for that experiment, you should go for it. And there are going to be families who are really willing for those experiments and to go to new schools. And we should thank them for taking those risks. And we should thank the schools for trying something new. And we should learn from them.

[01:31:02] And we shouldn't say that it's the perfect model. We also shouldn't say it's, you know, the most terrible model in the world. We should say what can we learn from this and how can other kids benefit from these folks who are willing to take this risk and be innovative. And experiment more. Yeah. Experiment more because it could – if you're spending more money, you could also take more risks. Some of them will pan out. Some won't. Let's learn from what there is to learn. I was, you know, having a one-on-one conversation with one of our senior most political leaders.

[01:31:29] And we're talking about the graduate income tax. And he said he's for it. And he wanted to explain to me why. It was because he thinks we need to spend more on education because our education system is broken. And I said, well, why do you think having less tax revenue is going to allow us to spend more on education? And he kind of paused for a second and said, well, if we raise the taxes through graduating income tax, won't we have more tax revenue?

[01:31:55] That's not what's happening in other states that raise it because what happens is those who are the high-tax, you know, payers, you know, migrate elsewhere. And they do steps in whatever planning they could do to make sure as much of what they would be exposed to is protected through other means. That's been the stats that have come out of the states that have done it. So some of these things that seem to be like an answer to a question of I really believe in education for, you know,

[01:32:23] for all of our kids, therefore we need to do the following could be the very thing that hurts our ability to build a better education for all of our kids. And driving away innovation from Colorado and the tax revenue that goes with that certainly is going to make it harder to, you know, you guys, and again, I'm not implying in any stretch that you've said anything along these lines, but if you decide as a result of that passing that you needed to relocate and with you goes those jobs and companies that would spawn off from you,

[01:32:53] what a loss that would be for Colorado, one that, you know, we would never make up. But that's the complicated decisions that we're facing. And when emotions get into it and not thorough enough thinking, we could be our own worst enemy. Yeah. Well, one, when we talk about learning and being a responsible citizen of the United States and being an agent and, you know, however you vote, however you do these things, AI is also an incredible enabler of you getting knowledge,

[01:33:22] asking questions and saying, hey, this is in the news right now. I watched a TV commercial. It said this. Can you give me the both sides of that? That didn't sound right. You can go to a chat bot and get a pretty thorough, thoughtful response of like, yeah, what he's saying is this. It's not 100 percent of the truth. Here's the thing. That's the story. Right? Like so when that's at your fingertips and that's so easy to get those responses and it does a pretty darn good job of researching,

[01:33:46] we have more responsible, informed citizens who will make choices at the ballot box that are better than the ones they made before. Like there's a reason why, you know, dictators burn books. There's a reason why they don't want folks to have knowledge. Right? Like this is the most instant, incredible knowledge building thing that's ever existed. Right? So it enables democracy.

[01:34:11] Including discoveries that will tackle head on our greatest challenges. You know, for example, energy. You know, we all want planet Earth that is our home to be as healthy as possible. We're not able to be as healthy as possible, yet we're consuming more energy. But it's AI itself that is already helping us discover new sources of friendly energy that will be so much more profound than anything we've discovered without it.

[01:34:40] And so whether it's in the case of, you know, you mentioned information. We've talked a lot about learning and teaching, discovering new forms of energy. I mean, society's going to change dramatically, hopefully mostly for the better. But there's a lot of very real threats and bad things that come out of AI as well. Which, you know, whether it's, if you could press a button and make AI not move as quickly, you know, I don't know. Maybe that would be a good button to press.

[01:35:09] But that's not a button that's really available. You can do that only in a very localized way. You can't do it across all of societies. And so if you do it in your little corner of the world, whether your little corner is New York or even all of the U.S., it's still going to happen. And it's going to happen in a dramatic way that we better do everything we can and prepare our kids to do everything they can to be on the right side of that equation of how to harness the good of AI and protect ourselves from the bad of it.

[01:35:38] And you, my friend, are doing exactly that in ways that the rest of us could talk about, but most of us aren't ever going to be able to do. And you are doing it. You are bringing kind of a platform to help educate our kids at scale across, you know, the entire world. You're, you know, you've made a lot of progress, but in a lot of ways you're just getting started in that journey. And I'm so, so happy to be able to see what you're doing and that you're doing it from here in Colorado and, and root you to success.

[01:36:06] And wish I knew you back when Range Ventures first knew you because I probably would have invested in you and that would have gone really well as well. Dan, thank you so much for having me. This has been a joy. And yeah, I think that there's so much opportunity here and I can't wait to see that the way that AI impacts education and the work that you're doing does as well. Great. Thank you. Thanks, Dan. Thank you for listening to this episode of The Bear Roars. Check out Stretch, the new song from Dan Caruso with music by Jason Mendelsohn, available now on all streaming services.

[01:36:36] If you enjoyed the episode, please like and subscribe on your listening platform. This podcast was produced by Loud Bear Productions and edited by Hannah Schmetzer with support from Kendall Weinberg, Alex Kim and Gibson Siegert. Thank you.