Taylor Lorenz’s Power User - Kids Are Being Taught By ChatGPT: Inside A $65K AI School

Episode Date: February 25, 2026

Would you pay $65,000 to have ChatGPT teach your kids?Support my independent journalism:🙏 Patreon: https://www.patreon.com/cw/taylorlorenz   🗞️ Substack: https://www.usermag.co   404 Media reporter Emmanuel Maiberg joins to discuss the disturbing truth behind Alpha School, an AI-powered private school that has been getting massive hype from mainstream media. While Alpha School promises parents that kids can complete their core education in just two hours a day, leaked internal documents reveal the very chaotic reality behind tech industry buzzwords.Alpha School charges up to $65,000 a year while openly aiming for a "no human in the loop" model to entirely eliminate human teachers. We discuss how their generative AI systems constantly hallucinate, creating illogical lesson plans and tests that treat kids like tech guinea pigs.Even more alarming is the massive data harvesting and surveillance. Alpha School utilizes software that monitors students' screens, web cameras, microphones, and mouse movements, acting essentially as boss-ware for children. We also expose how video recordings of students have been left unsecured in open Google Drive links and how the school's AI scrapes educational material from platforms like IXL and Albert.io, leading to disabled accounts and terms of service violations. Read the 404 Media article: https://www.404media.co/students-are-being-treated-like-guinea-pigs-inside-an-ai-powered-private-school/ Follow me:https://www.instagram.com/taylorlorenz      https://www.instagram.com/taylorlorenz3.0     https://www.tiktok.com/@taylorlorenz https://bsky.app/profile/taylorlorenz.bsky.social  https://twitter.com/taylorlorenz 

If you go to chat GPT right now, they're literally using that. So as a parent in many ways, you're just like, hear chat GPT teach my kid. And if that sounds like a bad idea to you, that's a lot of what is happening there in reality. Public education in the United States is a disaster. And as more parents are seeking alternative options, a new type of school has emerged that's been getting tons of hype. Alpha school is a new school with over half a dozen locations across the country that claims to be the first ever AI-powered private school where kids can get all their school work done in just two hours a day. But surprise, surprise, things are not exactly what they seem.

Emmanuel Mayberg, a reporter at 404 Media, recently got a hold of tons of leaked internal documents that reveal what's really going on at Alpha School. Today, Emmanuel joins me to talk about the rise of Alpha School, why its promise is so seductive to parents and mainstream media, outlets like the New York Times and how the school really operates. From invasive bossware surveillance tools to logically inconsistent AI generated lesson plans to the promise of eliminating all human teachers and more. We're going to dive into what's really going on at Alpha School and why all of this matters for the future of education. Emmanuel, welcome to Power User. Hey, thanks for having me. To start off, can you tell us a little bit about Alpha School

how it started and what the sort of original concept and promise was. So Alpha School advertises itself as an AI-powered private school. It has physical locations all over the country. It also has an at-home schooling solutions that parents can use to teach their kids. It really is originally something that was called two-hour learning, which was started by a woman called McKinsey Price. the philosophy behind that is we can get students to do all their core education in just two hours. Core education being all the studying you have to do in order to score well on standardized tests,

on your SATs, and things like that. And then the rest of the day can be used for more enriching education and activities, public speaking, team building, starting your own businesses, just for kids to pursue. whatever their passion is. At some point that morphs into the AI powered element of Alpha School, which is we can use Generative AI in order to make the learning that happens during those two hours more efficient and better and customized for each student. I was telling you this.

Like, I have seen it from Instagram Reels because I see these like Instagram hype reels, basically about the school and like how amazing it is. And it seems like from a lot of the media coverage online, it's received a lot of positive attention. Yeah. I mean, there's obviously, as you can tell by the expansion, a great desire for an alternative to traditional schooling and the one size fits all model that is what most people have. So it expanded very quickly. And I think not surprisingly, because it is very tech forward and has adopted AI really gained steam. with like Silicon Valley elites, the Trump administration famously Linda McMahon, the secretary of education under Trump went and visited one of the Texas campuses and said, this is the future

of education. So yeah, it's kind of like, it synergizes with like a lot of AI hype and tech hype and is popular for that reason. Yeah, I mean, I feel like it's being praised all over. I saw this NBC news segment, glazing them, box news. I also originally heard about it from the New York Times heart. Fork podcast where they had McKenzie on and talked about, you know, this exciting new model of education, which I'm not initially opposed to. Like, I think that the traditional education system is so broken.

And when you look at like just the decline of funding in public education in this country and the fact that like a lot of these old schools sort of public schools, they're falling apart. They're crumbling. People, you know, the kids inside of them don't have clean drinking water. They don't have clean air. they're huffing in, you know, disease all day and missing half the school year for being sick. And a lot of lesson plans probably are outdated and teachers are overworked. So, like, I feel like the concept of essentially scaling like this more like homeschool system using technology is an interesting one that I'm not initially opposed to. I think the generative AI of it all is what's concerning.

I was watching one of the videos where I think it was the NBC news segment where one of the kids has to, like, debate an AI for his final project, which it's funny because they're debating the topic of social media addiction, which is something I've covered extensively. And actually there's a lot of disputed research on. And it looks like the kid is like actually parroting misinformation even in this NBC news segment. So when did it get on your radar? And when do you start kind of like thinking, okay, wait, there might be more under the hood than, then, you know, it might initially seem. So first of all, I just want to agree with you. I have a two-year-old and the education system in this country is such that even at this early age, you have to start thinking ahead about where are you

going to send up to school because it's competitive at certain schools and people are very worried about certain public schools. And it's such a determining factor for the rest of your life, the kind of education that you get. So it was on my mind for that reason. And I also agree with you that the pitch is kind of incredible. Like the idea that a student can cram the rote studying and repetition of studying for something like the SAT and then use the rest of their time to pursue their passion, sounds amazing to me. Like, I wish I had that.

Whatever I was good at at school is the stuff that I cared about, like writing and everything else was like torture. So I totally agree with the pitch. And that conflicted a little bit with everything I know about generative AI, which is very faulty. And the idea that generative AI could entirely teach students on its own seemed ridiculous to me. And I was kind of poking around.

And eventually I heard from people at Alpha School and got access to some internal documents at the company. And it verified what I suspected, which is it doesn't work like they say. Yeah. Surprise, surprise. Tell me a little bit about the documents that you found and, you know, where some of the, I guess, reality doesn't match the hype. So conveniently for me, Alpha School uses the.

this program called Workflowy. I don't know if you've ever encountered it. Wait, they use Workflowie? Yes, they use Workflow, which I learned about via this article. It's like, I believe Notion is like the note-taking app that is also very popular among tech people. And this is kind of like that. It's a note-taking app. It's kind of obsolete. It was like a previous version like of what Notion ultimately became. Right. And it seems kind of clunky. And one of the reasons it seems clunky is that, you know, you have Alpha School. It's this big organization with people all over the world. And it's essentially like one giant scroll that contains notes from everybody. And like every employee goes in there and they talk about this is what I'm working on. These are the problems

that I'm facing. These are my thoughts about it. Here's what the challenges are. Here are some potential solutions. Here are some experiments that I ran about what could fix the problem. And that just told me a lot about how the AI actually works. And I think the best example of that is Alpha School heavily relies on this program called Alpha Read in order to teach reading comprehension all the way from grade school up to SAT prep. And one of the things it does is it can ingest a bunch of articles, then generate a text for a student to read, allegedly, again, according to their interests, and then produce questions about that text in order to check that the student actually internalize. the meaning of what they read. So this Alpha-Read AI system, it's taking large language models

from OpenAI, Google Anthropic, and then it's generating its own unique articles that these kids are then reading, which are AI-generated articles, and then they're being quizzed with AI-generated questions on the AI-generated articles. Is that right? Yes.

Okay. That seems like a mess. If you know anything about AI, your brain could also, could already jump to, like, why that would be concerning. And it's like, I don't know how well versed your listeners are in this, but it's like AI is prone to hallucinations. So you're going to get hallucinations, illogical questions, illogical texts in these programs.

Yeah, my friend was trying to use AI to study for the LSAT recently and sent me some screenshots where she was trying to do actually logic questions. And the AI just completely generated like false info. And finally she was like, wait, none of these make sense. None of this is logical. And, you know, ChatchipD, of course, is like, thanks. You're completely right. Thanks for putting that out.

She's like, oh, okay, well, I just did an entire AI generated diagnostic test that was, I guess, completely false. So, yeah, this is concerning. I mean, with this school, are there, like, tons of really well credentialed academic leaders around, you know, people with master's degrees in education who are, like, sort of reviewing this course material that are there to help the kids that are, like, you know, helping sort of shepherd this learning process. again, like using an actual background in education? So yes, but I think when I first started looking into this, Alpha School did have very well educated people at the company

who know aloud about AI and data and education. And they were there to develop these programs, but the ambition of the company is to get to a point where they're no longer needed. So increasingly, they are presenting students with materials that are not vetted by humans and subject matter expert to verify that the questions and lessons that they're taking are good. So they're there. A lot of them have left or have been fired. And what you definitely see is that students are actively getting their learning from completely automated AI systems.

If you're watching this video and you like my work, please support me on Patreon via the link below or buy a paid subscription to my tech and online culture newsletter at usermag.com. That's usermag.com. I don't have any long-term brand partnerships and a lot of my content is effectively demonetized. I've lost major brand deals for speaking out on certain issues and for challenging power. As you can imagine, advertisers are not exactly eager to work with somebody who covers a lot of the topics that I cover and talks about the things that I talk about. These videos I make are entirely funded by you, and I can't continue to make them without your support. So if you get any value out of the videos that I create and you want me to be able to create more, please support me on Patreon or Substack via the links below.

On Patreon, I do bonus episodes, monthly Q&A live streams, and post frequent updates about my work. My Substack newsletter gives you a bi-weekly roundup of everything that I'm seeing and reading and paying attention to online. You can also get my newsletter on Patreon. Once again, the links to everything are below in the description. Every dollar of your support makes such a difference. You write in your story that, you know, there's a lot of hallucinations. And, you know, in these internal docs that you got, there's these alpha school employees writing things like poorly constructed questions do more harm than good. They confuse students with unclear wording and illogical choices, undermining their trust in the assessment process.

These questions not only fail to meet SAT standards, but also fall short of the quality we promise to deliver. And then you also write. which I think is even more disturbing, is like these kids are asking for help, I guess, from a chat bot. And another Alpha School document reads, when a student requires help with additional questions, the chat bot fails to identify

which specific question is being addressed. And then you kind of go on and talks about like the accuracy of this AI tutor, et cetera. I should first say there are physical locations at the school, right? So if you're having a problem, you can raise your hand.

And there is a teacher there, what they call guides. who can come and talk to you and help you. And I think it's those people that are actually making the school function and making it so the kids know anything. But the guides aren't certified teachers, right? The guides are just there to kind of help the students as adults. But at least according to one interview, I saw, like,

they don't necessarily have a teaching background. That's true. And it's a mix. But the students also have access to like a network of remote tutors. So if you're working on some problem and you're having trouble, you can jump on a Zoom call and connect with a tutor who will help you. And that's how the actual teaching gets done, according to my sources. But any area of the company, any area of the education that relies heavily on AI, that's where the problems emerge. And there is more and more of that because the company is trying to push education entirely in that direction, right?

They call it fully autonomous education. No human in the loop. That's a quote. No human in the loop education system. And that's where you have the issue of AI generates a question. The question is faulty. They have an AI to check if the AI made a mistake, but that obviously doesn't work because it's like the same AI checking the AI's faulty questions

and not recognizing that they're faulty. Yeah. You write about this, like how actually they write these prompts for the AI to check the students work. And it's like, you're an experienced educational content evaluator tasked with grading a worked example, blah, blah, blah. You're an AI assistant tasked with creating a 20 question multiple choice. Like it just sounds very like shoddily sort of constructed. It's lazy. And again, to clarify, it's not like they're tapping into some secret AI model that they develop that is better than what you have access to.

if you go to chat GPT right now, they're literally using that. So as a parent in many ways, you're just like, here, chat GPT, teach my kid about, you know, English. And if that sounds like a bad idea to you, that's a lot of what is happening there in reality. Yeah, it's like not only are they, like,

learning off faulty AI stuff and getting hallucinated content, they're having to interact with the AI, which is like grading them, which also sounds terrifying and bad, I wouldn't want my grades for any school sort of thing determined by AI, unless it's like a multiple choice test or something. I'm sure, you know, the SATs are graded by AI or something like that. But to me, what stands out too about this like whole thing of like them having to hop on

Zoom is it just reminds me of so much stuff in the AI world that's promised to be autonomous and is not. It's like those like home robots that are actually remote controlled or like the little delivery robots that are remote controlled or like, I guess human beings were being used as well, to help with Waymo on the roads recently. But it seems like this whole thing is like, I mean, I don't think all AI is a house of cards, but I think there's a lot of promises being made that ultimately this AI can't deliver on.

And now we're back to relying on this like broken system of human beings this time through Zoom. Yeah. I mean, it's always the same story. We've been covering AI the entire time. And I'm with you. Like I am not a total on principle AI hater. There's clearly something there. There's clearly useful technology there.

The issue here is the same issue I've seen a million times, which there's opportunity to get investment. There's opportunity for a land grab. Alpha school thinks the future of teaching is AI, and they can be the first company there and be the dominant school in the country or the world. So they're all in an AI. We have to get there immediately. You get a lot of that rush and anxiety about this goal in the world.

workflowy and it's just rush and it's not ready. Like maybe it can get there one day. I don't know. I'm skeptical, but it clearly is not there now. And then, you know, to this other thing you said about it's not really a robot driving it around itself around LA delivering burritos. It's somebody in the Philippines driving the robot. It is the same thing here. It's like there's a quote in the story from a teacher who saw a bunch of kids get into like an AP biology class and they were like, these kids don't have basic skills to take this biology class, send them back through like the rest of their education and have a human verify that they can like read and write an essay. And then we can take them into the AP biology class. So it's like they went through a bunch of AI education. It didn't

really work. Then they had humans come in and fix it. It's so disturbing that parents are throwing their kids in this and it costs in some cases up to $65,000 a year. Like you're putting your kid in chat GPT school for $65,000 a year couldn't be me. Another thing that sort of is disturbing to me with AI and generally you guys do such a great job of covering like data privacy issues and privacy concerns. And you know, one thing with schools, it's sort of the incursion of technology into schools and, you know, all these kids using Google issued laptops in schools and having to rely on these like software programs for education is just that a lot of

those software programs are harvesting massive amounts of data. which they then go sort of like sell about the kids. And this is why I'm very against all of this stuff. And also I think why a lot of schools don't want kids using their personal cell phones, which they're increasingly banning, which Google loves because then the kids have to rely on their Google laptops, which they can then, you know, harvest data from. And I think in your article, you also talk about like Alpha School and the amount of data that they're collecting.

Is some of this also like a data play for them? It is a data play for them in the sense that, and they're transparent about this fact, all the data about what kids are. doing on their computer, what they're clicking, what websites they're going to, what apps they use, and then also the audio and video from their laptop cameras is useful to them because it's data, it's telemetry about how well their own programs for teaching are working. So they can see if a kid gets bored or if a kid gets up from their computer during some sort of exercise that they're doing. I think what parents did not realize is that the security and privacy around that data is not what you would hope.

And the most blatant example of that is all these video recordings of students are just stored in a Google Drive. And for anyone who has shared like a Google document via Google Drive knows you have kind of levels of permissions that you can give. It's like only this person at this email can access. this video or only this person with this type of email address that this organization can, or anyone with this link on the internet can see this. And that's how they are storing those videos on that latter option. So it's like if you have the link, if you used to work at the company, if you're at the company and you just send it to someone else, anyone has access to like hours and hours and hours

of kids doing these remote tutoring sessions. That's terrifying and very concerning. In October, as you mentioned in your piece, Wired reported that I-XL, which is an online learning platform that was used by Alpha School students, deactivated Alpha Schools account. They said that I-XL is no longer a customer. Alpha School violated their terms of service. You report that former Alpha School employees were like, we don't really know why the I-X-L account was deactivated. But Alpha School regularly uses other online learning platforms materials in ways that may violate their terms of service, either by copying or scraping. I was immediately thinking of like Khan Academy, which is also was sort of like early to this like online learning world and seems to have like actually a pretty robust like curriculum.

It seems like they're engaged in a lot of shady business practices. Yeah, like a lot of AI companies, Alpha School just indiscriminately and without permission scrapes whatever it wants from the internet. And there's two main reasons for that. One is they wholesale give students access to those materials in ways that they're not supposed to. And that appears to be what happened with I-Excel, right? So they buy one I-XL license, and then they tell the students here, go do these exercises.

That's not how the licensing system works with a platform like that. So I-XEL cut them off. Another thing that they do, and I notice because there is like extensive notes about this in the company documents is textbooks, platforms like Khan Academy, and others, they just take all that data and feed it to the AI models to create their own AI teaching products. And actually, I just heard this before we got on the call. I mentioned

another company, Albert I.O., which does something very similar to Khan Academy. There's notes in the document, in the company documents, that they terminated their relationship with Alpha School because it violated their terms of service. And somebody at the company just got back to me instead they did that because they saw that a bunch of employees at Trilogy, which is the company that develops all this software for Alpha School, was pinging their servers in a way, which made it clear that they're scraping their data, and they don't allow that.

So they actively terminated their account and gave them a refund to be like, we don't want to do business with you, don't take our data, we're done. I mean, it seems so unethical. Like it seems like those companies should sue alpha school. Like I don't know. It's like it doesn't seem like they did the work of even like you mentioned building their own curriculum. They're out here scraping.

I mean, it's just like AI mindset, I guess, where it's like let's just scrape whatever we need. We'll generate whatever. We don't have to worry about copyright. We don't have to worry about paying these companies that have invested years in building these educational materials. And we're just going to like test all of this on students. I mean, you talk about in your piece how like these students are essentially being treated as guinea pigs for this. whole experiment. Yeah, I heard that from all the former employees that I talked to.

It's that the tech is experimental. It's definitely not ready for prime time. And they're figuring it out on the go and the test subjects are the students. And I think what really broke my heart in a way is some of the data in the documents that I saw is feedback from students. And the students are frustrated. They know that they're being watched. They know that they're on this treadmill of just like completing as many exercises as possible.

And they can feel that. And I think that's really unfortunate. I want to get back to this like training data stuff and just the data harvesting element of everything. I mean, you mentioned trilogy. Do you think that this is all just sort of like some tech play? Like it seems like they're trying to build like a tech company that can get tech company valuation instead of a real educational system.

I mean, you mentioned in the piece how Alpha Schools work flowie lists ideas for enhanced tracking and monitoring of kids beyond screen time data. And you know, that the goal according to the note was to monitor the way kids are using apps and then use AI to analyze that activity, flag inappropriate behavior like bullying or drug use, and produce a general report about what kids are doing, you know, to like detect things like changes.

in friend groups or sentiment to flag emotional issues to parents and this just sounds like really scary mass surveillance type stuff even i mean i know you mentioned like they do catalog like how engaged a student is and you mentioned this app called study reel which monitors activity on their screens the computer camera the microphone you know how they're moving their mouse like it feels invasive like it seems wrong that we're doing all of this to children and harvesting that much information on these kids because as you mentioned i mean not just that google drive but who knows where this is information will ultimately go. Yeah, I think the thing about that type of data is I don't want to speculate about the intention behind it, but what I do know from years of reporting on tech is that once a pool of valuable data exists, it's sort of impossible to contain both because like that is the nature of computer security and it's like anything that exists in a system that is connected to the internet can be hacked. And then unfortunately, the value of of the data makes it so eventually it's sold off in some fashion or utilized internally for some product that you didn't initially know you agree to when you signed up right it's like we all signed up to facebook and we all started posting and like no one could imagine at the time that eventually it would be used to train lLMs and i think you can easily imagine the same thing happening with this kind of data there's also just a psychological effect to being monitored as any remote workers had a boss that uses this type of surveillance software and does Joe linement who's the principal at alpha school

and the co-founder of trilogy which is the tech company that's kind of seems in mesh in all this literally operates a functionally identical like surveillance software product called crossover that monitors remote workers and previous reporting describes as a software sweatshop and that the company's goal is to turn workers into algorithms and human CPUs like this just seems like this this like really invasive surveillance stuff and is clearly, I mean, we know the effect that this has on remote workers. And so to do that to children, like why is there not outrage? That's crazy. That's crazy that he literally runs like a software company that literally does this to workers and now is the principal at Alpha School and is maybe doing this to kids. Yeah, I mean, they don't pitch it this way,

but that's definitely how I see it. It's bossware for kids. And the way that one source described it to me is really all a company like Alpha School can do is present the learning materials to a student. And the thing that they are trying to do is to make their product better is enforce compliance, which is what Bossware does, right? It's like, I want to make sure that you're looking at the screen. I want to make sure that you're moving the mouse. I'm sure you remember from like early COVID, all these stories about mouse wigglers that remote workers had in order to like fool their bosses that they're actually at their

computers, we're just doing that for kids. It's like at the same time, there's a absolute like moral panic that I've been reporting on about like cell phones in schools, which again, I have to have the disclaimer every time that no one is saying kids should be able to sit and watch TikTok on their cell phones. But we know that's not actually how a lot of kids are using them. Some kids are using them as word processors. Some kids use them to communicate with parents, whatever, whatever. We can litigate that. But, you know, so much of removing cell phones from schools is is intertwined with surveillance of educational systems and it seems like Alpha school is kind of on the cutting edge of this type of surveillance like already

these classroom apps and Google and all these things like are harvesting data on public school kids and it seems like instead of Alpha school like taking a forward you know privacy-centric approach to tech and thinking maybe more forward thinking and like hey I'm gonna fix the problems in public education with you know AI and data harvesting and stuff it's like they're leaning into it yeah I mean that's how you would plan a few future for like an AI education company, right? You wouldn't want to cut yourself off from the extremely valuable data of how kids are using your products. You would want to be enmeshed

in their education. You want to see what they're doing, who they're talking to. This was like more of a brainstorming document, but you can see the thinking at the company. It's like they're not just interested in what the kids are doing on the apps. They're like, what is happening with them socially? Are they doing drugs? And it's like, it's a terrifying idea. And again, just an idea. But like, why is that even something that is being entertained at a company that is trying to educate kids? Right. And I mean, we're seeing these ideas entertained all over. We're seeing mass surveillance systems being enacted in public schools constantly. When they are pushing these cell phone bans, the number one reason parents are against them is because of school shootings.

And so a lot of these public schools are rolling out AI software surveillance systems in conjunction with the cell phone band saying, don't worry, we have AI scanning all the kids. It can identify a gun. It can identify, you know, deviant behaviors. And we saw what just happened recently, I think it was in Baltimore where a kid was tackled to the ground by the police because he had a Doritos bag

that the AI identified, you know, misidentified as a gun and, you know, has trauma from that. Like, it just, all of these systems seems really broken. And I'm wondering why there is an outrage, you know, at this time when there is this like intense obsession with like child online safety and we have to protect the children. and all of this stuff, like, you see nonstop. I mean, even in that NBC news segment,

at the same time that they're pushing this idea that social media is addicting, we have to protect the children, and that's what one of these Alpha Kids is doing, you know, a very misleading presentation on. Also, we need to roll out mass surveillance in, like, AI school. Like, it just seems so incongruous. Why aren't there more people sort of protesting this type of stuff and protesting, you know, what Alpha School,

that sort of that vision that Alpha School is putting forward? I think I have an answer, but before the answer, It just reminds me the CEO of Axon, which is like the body cam company at some point was pitching that school have this device attached to the ceiling where once it detect that there's an active shooter or something, it deploys a drone that then flies out and, you know, shocks them into submission. You could just imagine that being married to like some sort of like alpha school type surveillance where it's like, this student is slow in his math question. and just like zap him or something. But in terms of like why this is happening, I think something that I've learned from covering AI for a few years now is it's a lot of snake oil. And looking at it for like two minutes critically will make you realize that.

But wherever generative AI finds a foothold, it's a place in society or business or culture where the existing system has already failed. Right. So as we said, when it comes to education, people are really desperate for something else. And that is not made up and that's not a scam. It's like people really have a desire to have education alternative. And if you have the money and you have the means and somebody says, hey, this AI school will give

you better test results and a happier kid and all this stuff. It's like, yeah, of course you would try that. Like, why wouldn't you? Yeah. I mean, I think of this. Like, I had such a hard time in public school. This is why I don't want to send my kids to public schools. because like and I report on how broken some of these schools are.

And you know, unless you live in a really fancy area and you're really rich and you have really good public schools, sure, then I'd love to send my kid to a really nice public school. Most public schools in America are not like this. So I just, I know a lot of parents increasingly too that are doing these homeschooling pods. And AI seems to be something that they're very interested in because like you said, it's being sold as this thing. And listen, sometimes it can be a useful study tool. I'm not going to say that like chaty BT learning mode is like useless. I'm sure some people get some things out of it, but I don't think it's like ready for the big stage

in the way that these people want it to be. And the fact that there, that the alpha school people too, like, it's not even like Chachy PT is doing this. It's like these people that didn't even make their own AI LLM, like didn't create any of their own curriculum material. Like don't even have a lot of like professional teachers in these classrooms and are like trying to get kids to just like, I don't know, Zoom, you know, with others when they have questions.

Like it just seems like a complete disaster. And they're charging tens of thousands of dollars for it. That is so obscene. And I feel bad for these parents. Although I will say, and you note this as well, like, I don't know the test scores are okay. Like a lot of these kids do seem to be doing pretty well. How do you explain that? This is something that I'm happy to say McKinsey Price, who founded this company will readily

admit there is a selection bias happening. If you have $65,000 to send your kid to a price. private school and you care deeply enough about their education to do it, that might explain why they are getting good grades more than anything else that is happening at the school. And there's mountains of research that shows that income is positively correlated with good grades. So that is most likely what is happening. I will also say, sometimes I cover a company, sometimes it's an AI company, sometimes it's a different company, and I walk away from the experience thinking, like, this entire project is irredeemable and very bad. And that's not my take on Alpha School after having covered it pretty deeply.

I think there is a false promise there about AI. I think it's implemented badly. I think the people I talk to who work there will say it. But what they also said is that they love these kids. They really care about education. They really want them to succeed. And they put in the work to make them succeed despite the false promise of AI. And the grades are good.

So that is not something that I will try to deny it. Just that when they say no human in the loop is their goal, what they really want to say is like, we don't want to pay teachers ever again. We want to take $65,000 and spend it on zero humans that will teach your kids. and that's where they're going. And that, I will say, is, like, a bad ambition to have. Yeah. I mean, it seems part of this, like, reactionary movement that is very intertwined with the rise of AI, which is about, like, eliminating workers and eliminating jobs

and alpha schools proof that you do need human beings to kind of be there to support kids as they learn. I'm totally with you in the sense that, like, I'm willing to accept, even that there will be, like, AI sort of infused education. Like, these feedback, I think it needs to be, way more tested and like over there's a lot more oversight and all that but like I don't know I'm a believer in technology we are tech reporters we love technology you know theoretically otherwise why would we write about it all day but like yeah it seems like the the marketing is kind of where it's gone

wrong and the surveillance aspect to me is inexcusable absolutely agree agree entirely well emmanuel thank you so much for joining me today thank you for having me all right that's it for the show if you like my work please please support me on patreon via the link below. You can also buy a paid subscription to my substack newsletter at usermag.co. That's usermag.com. I send a biweekly roundup of everything that I'm reading, seeing, and following online. You can also get that newsletter on my Patreon where I do bonus content, a monthly Q&A, live stream, and more. As you guys probably know, I produce and fund this podcast entirely on my own. So seriously, every single dollar of your support makes such a difference,

and I cannot thank you enough. The link to my Patreon and substack are below. Thanks, and I'll see you next for a brand new episode of Power User.