Summary
Cognition had gone up every single generation since the late 1800s. Gen Z is the first to go down, and they spent more time in school than we did.
Dr. Jared Cooney Horvath is a teacher turned neuroscientist who studies how humans learn. In January 2026 he testified before the US Senate that pulling smartphones out of schools is only step one, because the laptops do not teach either. His book The Digital Delusion, re-released by Harmony on August 18, 2026, lays out the data showing that student outcomes fell as technology entered education, and then explains the mechanism underneath it.
The mechanism is transfer. Learning only moves in one direction, hard to easy. You learn to handwrite before you type, and to do arithmetic before you reach for a calculator. Technology exists to strip friction out, so a child who starts easy has nowhere harder to go. That produces the distinction at the center of this conversation: an expert who hands a task to a tool is offloading, because they can vet the output the second it appears. A novice doing the same thing is outsourcing, and learns nothing at all.
Host Ryan Vet, who does not allow screens in his own house, presses on where that leaves parents, schools, and AI. Horvath's answer on AI in K through 12 is unambiguous, and his answer on whether Gen Z can recover is two words: heck no, it is not too late.
Key takeaways
1
Learning only transfers in one direction, hard to easy. A child who learns to handwrite can move to typing, but a child who starts on the easy tool cannot move that skill into anything harder.
2
An expert who hands a task to AI is offloading, because their expertise lets them vet the output immediately. A novice who hands over the same task is outsourcing, and learns no facts and no skills in the process.
3
Almost all of the measured benefit of an AI tutor shows up on the tutor itself. When researchers test whether students can apply the skill off the program, scores fall sharply, the same pattern that ended the brain training industry.
4
Learning from a human teacher produces physiological synchrony, with an effect size of 0.68. There is no biology in a screen for a child to resonate with, so an entire mechanism of human learning is simply removed.
5
Between 2013 and 2023 the share of kids worldwide meeting even minimum digital literacy dropped 13%, while the share explicitly taught to use computers at school rose by hundreds of percent.
6
PISA first measured creative thinking in 2022. Singapore ranked first at 41 out of 60 against an OECD average of 33. Most systems that scored well in creative thinking also scored well in math, reading and science, though several did not.
7
Seven months after the book first published, proposed state ed tech bills went from 3 to 31. Horvath calls the fix a reversion rather than a change, because the analog infrastructure still exists.
Terms defined
Plain-language definitions for the ideas in this episode. Structured for search and AI answers.
transfer noun · learning science
The movement of knowledge or a skill from the context where it was learned into a different context. It reliably runs hard to easy, from handwriting to typing or from arithmetic to a calculator, and it does not run in reverse.
In this episode: Horvath's answer for the single biggest reason classroom technology fails.
offloading verb · learning science
Handing a task to a tool that you already know how to do yourself. It works because your expertise lets you judge the output the moment it appears.
In this episode: What Horvath does when he asks AI to run statistics he could run by hand.
outsourcing verb · learning science
Handing a task to a tool with no knowledge of the process and no ability to judge whether the result is right. Same tool as offloading, opposite outcome.
In this episode: What happens when a novice is given an expert's productivity tool, illustrated by the rented e-bike that runs out of battery.
physiological synchrony noun · neuroscience
The way two people in genuine rapport begin to match physically, with heart rates, breathing, and blinking falling into step and brain activity converging. The formal term for the underlying process is affective empathy.
In this episode: Why a human teacher has a 0.68 effect size on learning and a screen has nothing to resonate with.
cognitive erosion noun · Ryan Vet framework
The gradual loss of a human capability as each new tool removes the friction that built it, from spell check flagging a word to Grammarly rewriting the sentence.
In this episode: Ryan's frame for the slow handoff of thinking to software, which Horvath grounds in the data on transfer.
Notable quotes
Lines worth keeping, verbatim from the episode.
They won't learn facts, they won't learn skills, they'll just learn dependence on the tool.
It takes human beings a couple years to learn how to speak, but it takes them decades before they have anything meaningful to say.
Your generation uses technology. My generation invented technology, which means it represents our way of thinking way more than it represents yours.
No one asked to have a screen put in their face for six hours a day every day at school.
School is nothing other than a giant ban. It is a recognition that kids are not ready for the real world.
As long as you're alive and breathing, this whole system is still malleable. So keep pushing it.
If they can learn the human skills of negotiating the playground, of working with a classmate, or even interfacing with a teacher, then they'll be fine prompting a machine.
Chapters
Jump to any moment. Timestamps deep-link the audio.
The guest
JC Headshot pending
Jared Cooney Horvath
Educational neuroscientist · Director, LME Global
Jared Cooney Horvath is a teacher turned neuroscientist. He started in the classroom teaching grades 6 through 12, went to study the brain during what he calls the decade of the brain so he could bring the science back to his students, and has now spent about 16 years in academia. He earned a master's in mind, brain, and education from Harvard, did a research fellowship at Harvard Medical School under Alvaro Pascual-Leone, and completed a PhD in cognitive neuroscience at the University of Melbourne. He lectured at the Melbourne Graduate School of Education from 2013 to 2024 and remains an honorary researcher at Melbourne and at St Vincent's Hospital Melbourne. He is the director of LME Global and the creator of The Learning Blueprint.
He built his reputation debunking overclaims rather than making them. His most-cited work found no reliable cognitive benefit from transcranial brain stimulation, directly against the commercial hype of the mid 2010s, and he has spent years pushing back on pop neuroscience that outruns its evidence. He now applies the same standard to educational technology. His argument is not that screens are distracting. It is that student-facing, internet-connected devices conflict with the mechanics of learning itself, which he calls a category problem rather than a software problem. He describes himself as pro-rigor, not anti-tech, and draws a hard line between teacher-facing technology, which he considers largely fine, and one-to-one student devices, which he does not.
His book The Digital Delusion was self-published in December 2025 and is re-released by Harmony, an imprint of Penguin Random House, on August 18, 2026, with roughly 25% more material. In January 2026 he testified before the US Senate Commerce Committee on technology in schools, arguing that removing smartphones is only the first step. He writes the Substack The Digital Delusion and posts learning research to his YouTube channel.
Frequently asked
Straight answers to the questions this essay raises. Open by default, structured for search and AI.
Why does Jared Cooney Horvath say classroom technology harms learning? +
Because of transfer. Learning only travels in one direction, hard to easy. You learn to handwrite before you type and to do arithmetic before you use a calculator. Technology is designed to remove friction and keep things easy, so a child who starts easy cannot move that skill into anything harder. They learn dependence on the tool instead of the skill.
What is the difference between offloading and outsourcing? +
Offloading is what an expert does: they hand a task to a tool because they already know how to do it and can immediately vet the output. Outsourcing is what a novice does: they use the tool with no knowledge of the process and no ability to judge whether the result is right. Same tool, opposite outcome. Horvath's example is a rented e-bike that runs out of battery. The fit rider pedals home. The unfit rider calls an Uber.
Do AI tutors actually work? +
Horvath argues they work only inside narrow, right-or-wrong domains, and that almost all of the measured benefit shows up on the tutor itself. When researchers test whether students can apply the skill away from the program, scores fall sharply. He compares it to brain training apps, where players got better at the game and not at anything else.
Does Jared Cooney Horvath think AI belongs in schools? +
No. He argues AI has no place in K-12 because it is not a learning tool, and that school is by design a set of bans that shields children from the adult world until they are ready for it. He does distinguish teacher-facing technology, which he considers largely fine, from student-facing internet-connected devices, which he says are where learning collapses.
Is it too late for Gen Z? +
No. Horvath's position is that the brain stays malleable, so the capability can be rebuilt with effort. His hope is that as Gen Z recognizes what was done to them, they push back and demand something different for their own children.
Resources mentioned
Books
The Digital Delusion See details →
Get the book
Also mentioned
Newsletter The Digital Delusion on Substack → YouTube Jared Cooney Horvath on YouTube → Organization LME Global → Framework Cognitive erosion → Framework The friction doctrine → Episode Joseph Ferris: Boopsy, Gen X, the Barbell Generation, and Embracing the Suck → Episode Tim Estes: AI and the Internet Kids Deserve → Parenting & childhood Raising the Next Generation: restoring friction →
Transcript
00:01 Cold open
Jared Cooney Horvath · On this episode of the Ryan Vet Show. Every generation outperforms their parents on everything. Until Gen Z rolls along. They spend more time in schools than we do, yet all of a sudden their cognition is going down for the first time in recorded history.
Ryan Vet · Is it too late for Gen Z?
Jared Cooney Horvath · Heck no!
Ryan Vet · Every single generation has certain friction elements stripped out of it.
Jared Cooney Horvath · They'll just learn dependence on the tool.
Jared Cooney Horvath · When his bike ran out of juice, he had to call an Uber to take him home because he was outsourcing.
Ryan Vet · If they can learn the human skills of negotiating the playground, of working with a classmate, or even interfacing with a teacher, then they'll be fine prompting a machine.
Jared Cooney Horvath · As long as you're alive and breathing, this whole system is still malleable. So keep pushing it.
00:41 Meet Dr. Jared Cooney Horvath
Ryan Vet · Welcome to another episode of the Ryan Vet Show. I'm excited to have with me today Dr. Jared Cooney Horvath. He is an author, and I actually first heard of him from C-SPAN. Now, I can't say I spend my days watching C-SPAN. That is not what I do. Only my weekends. No, just kidding. I don't really watch it that often. But he had a line in there, and it was contrarian to a lot of popular culture, but something I really agree with, and it was this idea that youth and kids today are just not as cognitively capable as generations before. He's got a new book coming out, The Digital Delusion, that is coming out in just a couple weeks. So I want to talk about all that. But before we get too far into it, Jared, I'd love for you to share your background and share your story.
Jared Cooney Horvath · Yeah, absolutely. And thank you so much for letting me be here. So I am a teacher turned neuroscientist. Teaching was all I cared about. But I was teaching back during the decade of the brain. So I just remember we had all this brain jargon going around, brain books, brain gym, brain this, brain that. And I thought that all sounded very cool. Why don't I go learn that so I can bring that back to my classroom and say, here's how this all works? And so I genuinely thought that'd be like a year or two of my life, but I've now been stuck in academia for about 16 years. And I study now basically anything that has to do with human learning. So neuroscience, psych, AI, behavioral economics, you name it. If it deals with learning, my job is now to synthesize all that, and I get to bring that back to teachers and students and communities and say, this is learning. And if this is learning, what does that mean for teaching? What does that mean for studying? What can we do at home? So I have a very fun job now.
02:18 A Senate hearing, and 25% more book
Ryan Vet · I would say so. That sounds great. And I know one of the things that you do, you often debunk some pop neuroscience things that might overclaim just where we are in culture and society, because sometimes it's fun to have great studies that make great headlines, but you love to get at the root of it. And I think that's some of what your new book coming out, The Digital Delusion, is looking at. Could you talk about, it's actually a re-release, if I understand correct, is that right?
Jared Cooney Horvath · Yeah, yeah. This will be version two with 25% more book, we say. Just to add material.
Ryan Vet · I love it. Well, could you talk about the journey with the book, and then what to expect in this new 25%?
Jared Cooney Horvath · Yeah. So, kind of like what you heard on C-SPAN, which is funny because I'm same as you, I never watched C-SPAN, oddly enough. But I was asked to testify before the Senate. They were doing a lot of work on smartphones in schools. And so I came in to kind of be the voice about learning, because I think there's this mistaken belief that smartphones are the danger. If we get rid of those, at least kids can still use computers to learn, and they have their laptop and they'll be good. And I come in and go, no, no, no, we don't learn from these things either. So getting rid of the smartphones is step one. We now have to have a real think about the laptops and the rest of the tech we are doing in school. So what this book lays out is basically all the data that shows that when tech entered education, learning goes down. And that goes back to the first point you were making. If we look at Gen Z, since we've been recording cognitive abilities since the late 1800s, every generation outperforms their parents on everything. Literacy, numeracy, working memory, executive function, IQ, you name it. We always go up. And the reason for that was school. A lot of these cognitive measures are tied directly to school. School is where we build our cognition. Each generation spent more time than their parents in school, therefore our cognition goes up. So it was always this kind of one-to-one shot, until Gen Z rolls along. They spend more time in schools than we do, yet all of a sudden their cognition is going down for the first time in recorded history. So something has decoupled cognitive development from school. And our big argument, especially in this book, is that it was the tech. It was when we introduced computers and when we introduced digital learning and digital tutors and Khan Academy and all these videos and stuff. We just don't learn that way. And so that's what made us start to learn worse and our cognition start to go down. So the book lays out the data and the mechanisms. My goal was then to explain why. Can I tell you why this is happening? If so, now we have good reason to start pushing back and say, all right, what do we have to do to make school better?
05:04 Transfer: learning only travels hard to easy
Ryan Vet · I love that. And I think one of the things that I argue and have argued for many years is every single generation has certain friction elements stripped out of it, meaning we've tried to make it easier for the next generation in every category of life. We've tried to make learning easier or more efficient or more accessible. And with the increase of the access to information and the access to quote unquote knowledge, I feel like we're losing some of that experience and wisdom and the friction that comes from learning, and technology is a major catalyst. If you had one major, and I don't want to spoil the book, I want everyone to get the book, but if you had one major aha moment, what would you say is the key? This is why we're seeing technology not work like we intended or thought it would.
Jared Cooney Horvath · I'd say the biggest lever of all this is going to be a concept called transfer. And the idea of transfer is, how do we learn knowledge and skills in one context and then apply them to another context? So, like, I used to teach math. I'm glad my kids can do math in my classroom. I need you to be able to do math at home when you're balancing your family budget or whatever it's going to be. So how do we move knowledge and skills? Now, the big trick is this. When it comes to learning, it has always been hard to easy. Learn the hardest stuff first, and then you transfer that into easier and easier and easier settings. So all the hard stuff starts basically at the beginning. Learning to handwrite is one of the hardest things you could possibly do. Learning to do arithmetic physically is incredibly difficult. But as you get good at these, then we can move into typing, then we can move into using a calculator, now we can move into voice to text. Start hard, move easy. When you try and transfer the other direction, start easy, move hard, it doesn't work. Basically, if you start by learning super easy, you cannot move it into anything more difficult than where you learned it. So you're stuck with just the easiest stuff you ever learned. So now, why is that tricky for tech? No one invented tech to make our lives harder. Right? No one invented a computer to make life harder. Like, here's a GPS that's harder to read than a real map. The entire purpose of this is to make everything as smooth and easy and frictionless as possible. Which, as an adult who's already learned everything, that's great. It saves me a whole heck of a lot of time. It's a production tool. But for a kid who's just learning how to become an adult, that kind of ease and frictionless living, that is the surest way to make sure they will not learn anything. They won't learn facts, they won't learn skills, they'll just learn dependence on the tool. So I think transfer becomes the biggest lever for all of this discussion.
07:47 Cognitive erosion, from spell check to Grammarly
Ryan Vet · That's so fascinating. And you can kind of look at cognitive erosion over time. We had spell check in the early days, and it would basically underline something red and say, this is not in the dictionary. And that's all it said. It wouldn't even give you suggestions or anything like that. And then slowly you had the green lines next in Microsoft Word back in the late 90s. And the green line meant there could be a grammar error here, but we're not going to tell you what it is or how to fix it. Then it started with more prescriptive analytics telling you, hey, this is what you need to do. And all of a sudden we've got Grammarly now, where I click a button and my sentence is reworked, and I didn't write it anymore. And I think we're starting to see this cognitive erosion even with simple things. It's embarrassing, but I'll admit it, I had to learn how many cups in a pint and pints in a quart and quarts in a gallon. I think I said that right. My fourth grade teacher is going to be mad, or first grade, who knows? But now we have our robot friends sitting on our kitchen counter and we can say, hey, how many teaspoons are in this again? You said something really interesting I want to come back to. You said as an adult, you've already learned the hard things, or you've learned some of this. And so now you get to enjoy the convenience. But for kids, we seem to be bringing them into our current convenience as adults, and I think that's causing a big miss for many kids being raised today. What are some of those points of friction that we could put back into kids, not to exasperate them, but to actually prepare them?
Jared Cooney Horvath · Everything. There's very little that we've taken away that I would argue has been a good removal of friction. By all means, if we just go back to learning, hard copy books, taking notes by pen and paper, reading long-form text as opposed to just watching videos, having to do physical projects, having to work with other people and actually present what you've learned. All of these things are how we learn. And that's the point. We've mistaken adult productivity for student learning. And a very good rule of thumb: the tools you use as an adult to make your life easier are never the tools you should give a novice in order to help them learn how to become an adult or an expert. So probably the best example, and you were dancing around it, is AI. That's just now come along and said, we'll do everything for you, don't worry about it. Now, AI, if it can be said to do anything, and I still don't even believe this, you can say AI is an expert offloading tool. And when we say offloading, what we mean is, as an expert, I already know how to do something, so I'm going to let the tool do it only because my expertise can allow me to make it meaningful. So, for instance, I can do stats, right? But it would take me hours to do stats by hand. So what do I do? I ask AI to do it, it'll do it in two seconds. But it only works because I can immediately vet the output. As soon as it spits something out, I can tell you, yep, that's right, or no, that's wrong, I have to do it myself, or something's typed in there wrong. My expertise is what allows it to all function. Now, when you give that same tool to a novice, they don't do what's called offloading, they do what's called outsourcing. They use the tool with no knowledge of what's going on, what process is happening, whether the output's good or bad. So all they learn to do is type in a prompt, copy-paste the output, hand it to someone, and see if they smile. In the interim, they don't learn any facts, they don't learn any skills, they don't learn anything. So that's the difference. The exact same tool in one context can be offloading, and in a different context can be outsourcing, which is highly detrimental. And just to clarify, the example I always use is my sister-in-law is really fit. She's a gym rat, and her husband is really large. He's not a gym rat, he does not work out. And so they rented e-bikes to go tour around their city. As always happens, the e-bikes ran out of batteries. In her case, it didn't matter. Why? Because she's fit. She was just offloading, she was able to pedal that thing right back home. In his case, when his bike ran out of juice, he had to call an Uber to take him home because he was outsourcing. He didn't have any skill, he didn't learn any skill, he was just using a tool to get something done. And when the tool collapsed, he had no recourse. He didn't know what to do to survive it. So that's why you're right. We cannot mistake what we are doing as adults for what kids need while learning how to become an adult. Very different ballgame.
12:01 Do AI tutors actually work?
Ryan Vet · I love that analogy of the e-bike. That's fantastic. You had something on Substack recently, and subscribe to Jared's Substack if you're not already there. You had a multi-part series on AI tutors. Because on one hand, there's this great opportunity, because you can learn anything online and now it can adapt to the places where you're weak. And that's the argument that a lot of people make. That's not the argument you make. Could you talk a little bit about why AI tutors might not be the saving grace that we think that they might be?
Jared Cooney Horvath · Yeah, I think that's going to bring us back to transfer, oddly enough. So the data on digital tutors is one of the only places where it appears that ed tech can work. Most examples of ed tech just totally underperform, but when it comes to digital tutoring, it looks okay. The caveat to that is, one, they will only work in highly constrained fields, typically fields with right or wrong answers. So we call this surface-based learning. It can teach you how to do math, but it can't teach you why you would ever want to do math, or it can't teach you applied calculus. It can teach you how to read, but it can't really teach you what reading is for, reading for learning. So we always say that you can learn how to read on a digital tutor, but you can't read for learning on a digital tutor, because now you're going deeper and computers just can't handle that. But the second big caveat is transfer. All of the impact of digital tutors, or the vast majority of it, comes on the tutor itself. So basically the data is saying you are getting better at the tutor. Well, cool, man. I bet that's fantastic. I need to know if you're getting better outside of the tutor. And that's where all the data really starts to falter. You see, when you take a look at, can they move the skills off of the program, scores plummet significantly off of that. So AI tutors feel good, but it's just like brain training. Do you remember when brain training was a thing? We had Cogmed and Lumosity, and that became a billion-dollar industry. And today you don't hear about it, it disappeared. Why? Because they found out through all the data we collected. When you do a brain training thing, you get better at the game, but you do not get better at anything else in the real world. So you end up just basically wasting your time. So it becomes a measurement issue. So, yeah, believe it or not, AI tutors: if you want your kid to be really good at an AI tutor, go for it. If you want your kid to be able to apply that in the real world and actually understand what's going on around him or her, probably not the right way to do it. That's where human teachers come back to the fore.
14:33 Synchrony: what a human teacher does that a screen cannot
Ryan Vet · And I think there's so much more to a human teacher, too, because you can disappoint the human teacher. You can see that. You can see when you bring them joy, when you've done something well. You can have friction with your teacher and conflict with your teacher. And I think that's just something that AI can't provide. It's the streamlined, frictionless world in which, yeah, you might learn something, but it's rote memorization, and in the real world, throw you in a different scenario, you're up a creek without a paddle.
Jared Cooney Horvath · And here's another interesting mechanism, then, there too. As human beings, our biology can do what's called physiological synchrony. It can resonate. And the official term for that is affective empathy. So if you and I are empathizing right now, our hearts are going to start to beat at the same pace. We're going to start to breathe in synchrony, we're going to blink at the same time, our brains will start to look almost identical. It turns out that synchrony becomes one of the strongest drivers of learning. When a kid is syncing up with a human teacher, for the nerds out there, it has a positive effect size of 0.68, which in education is huge. It's exactly what we want. But realistically, what's going on is, in that moment, the kid is no longer learning from a person. They are thinking like that person, which makes understanding so much easier. I'm making all the same cognitive moves you are. So in a very real sense, what this shows is human beings have evolved to learn from other human beings. We have built-in mechanisms that say, if you see another human, learn from them. This is what we do. You bring that same learning online with an AI digital tutor, there is no biology to resonate with. Literally, an entire aspect of human learning just gets tossed. So this is why we tend to say, and to clarify, when it comes to digital learning, you can learn. It's not that we don't, it's just that that learning is much slower, much shallower, much less transferable than the exact same learning you would have done even with just a human being in the room with you. So that's another biological mechanism that says, if you get rid of that, don't be surprised when learning goes down.
16:32 Raising kids off screens, and prompting a machine
Ryan Vet · That's really profound. And one of the things, you know, I grew up with an Apple II and a Mac Classic, like old, old, OG, not connected to the internet, floppy disk drive. You had to type MS-DOS code to get the dinosaur to walk across the screen. And so computers in that era, I lived fully outside and fully on the computer. I grew up in the perfect era where I had just enough technology exposure, and it allowed me to do a lot of things that some of my peers didn't even get to do. And so everyone asks me, I've got kids, my oldest is going to be a first grader, and it's, hey, what are you allowing? Are you letting him use AI? Are you letting him use computers? And I tell them no. We don't do screen time in our house. They don't know how to use my phone. They aren't on AI. And everyone thinks that I'm absolutely out of my mind for that. And I said, hey, I've been working in the AI field, or at least predictive analytics and prescriptive analytics and machine learning, since 2010 in the venture-backed capital days when I was in that space. And everyone asks why. And I say, and I'd love your take on this, I say, if they can learn the human skills of negotiating the playground, if they can learn the skills of working with a classmate or even interfacing with a teacher, then they'll be fine prompting a machine. What are your thoughts on that? Am I thinking right, or am I totally off base?
Jared Cooney Horvath · 100%. Oh no, you're 100% correct. This whole argument that kids need tech to succeed in a tech future job market is absolute nonsense. In fact, the data is very clearly going the other direction on that one. Between 2013 and 2023, in that 10-year window, the percentage of kids around the world who meet even minimal digital literacy, what you're going to call the thresholding, has dropped 13%, even though the percentage of kids who are explicitly taught how to use computers at school has increased hundreds of percent. So what you see is a very big disconnect between learning the tool and using the tool well. Why? Because if you want to take away one thing from 17 years of learning about learning, it takes human beings a couple years to learn how to speak, but it takes them decades before they have anything meaningful to say. That's why education has never been a tool-based system. Learning how to use a tool is easy. AI is a freaking box you type words into. How long do you need for that? I can teach you that in about two minutes, it's fine. Surely the thing that matters is what words are you typing in there? And I don't mean prompt engineering, I mean what are you bringing to this tool that is going to make it meaningful. Hand me a hammer, I know how to use it. I don't know how to build a wall, though. So I can use your tool, but I've got nothing to do with it. Give it to a carpenter, he can use it exactly the same as I can, but he can actually build stuff with it. It wasn't the tool, it was the knowledge underlying the tool that mattered. That's why school, K through 12, has always been a general education. We are not teaching you history because you're going to become a historian, or writing because you're going to become a novelist. We're teaching you how to think through knowledge so that when it comes time to start producing, you have deep understanding of learning and processes that you can apply to a tool. And honest to goodness, it would take six months. This is what work used to do. They would hire people and then train them how to do the job. Instead, they now hire people assuming they know how to do the job, and then are shocked when they don't know anything else. Well, it's because you stole all their other time teaching them how to do a tool, which you could have done in six months. Instead, you took 12 years of their life to pound tool use into their head. So, no, by all means. Millennials are more digitally literate than Gen Z, and we were never taught how to use digital technology. That should be the only example you need, that if you teach human beings how to think and learn, they can adapt to whatever tool comes their way. But if you teach kids on today's tools, they will forever be stuck on today's tools. And when the world moves on, they've got nothing to push beyond that.
20:35 Why millennials are more digitally literate than Gen Z
Ryan Vet · I think that's really interesting. And I've seen many studies on this where millennials are more technology literate than Gen Z, and they're also the ones pushing AI faster than any other generation. And this is a generation who now you're looking at, they're in their 30s and 40s. They're not the kids these days that so many corporate people get frustrated at. And what I find really interesting, millennials had a couple things going. They still had some of the benefits of the Gen X lifestyle, being outside, enjoying real life, but they also evolved with technology. They don't remember internet being wireless when they were kids. That was something that happened as they were older. And every time they logged into the Facebook, it changed from the Facebook to Facebook.com, then your wall turned into your timeline and the post. So they were evolving as technology was evolving. But if you look, there's kind of this plateau in a lot of the evolution of some of these systems in the late 2010s that flattened out, and all of a sudden that's the world Gen Z inherited, and it hasn't changed. So they not only developmentally growing up didn't have their critical skills, but they haven't been in a technologically evolving world like millennials were.
Jared Cooney Horvath · Yeah. And so what do we do? We drill them on one thing. Do you know how to code, quote unquote? That's about the only thing you can think of. And then we send them into the real world where everyone can code. It's irrelevant. It's what are you coding? What do you have to say? What have you learned that you want to bring into this world so you can make it a better or a different place? And they just never really locked any of that down. And that's the other idea also, of the calcification of culture. Look at movies from the 70s versus the 80s. You'll know exactly what you're looking at. Listen to music from the 90s versus the 60s. It's not even hard. Watch a movie from 2005. Watch the first Avatar and then the last one. They look exactly the same. They feel exactly the same, even though there's been 25 years. Why? We've calcified. When everything becomes personalized, everything becomes basically unified onto what the tool is going to allow you to personalize for yourself. And then we stop creating new trends, we just start chasing common trends to get the most use out of the tool. And so yeah, I feel bad for Gen Z, but they've lived in a very flat, stagnant culture for a long time. But my biggest hope for them is that they recognize that none of this is their fault. When we say they're cognitively less than us, when we say they don't have the same identity markers that we had, no one asked to have a screen put in their face for six hours a day every day at school. They didn't ask us to do that. We did that to them. So my hope is that when they hear these kinds of conversations, and as they get older and they start to realize that they're kind of lacking some of the basic abilities to just navigate the world, that they get mad. That they get mad at us and they start to push back and they demand different for their kids and for their families than we did to them. I think we made the big, you're right, millennials, we were the tech ones. Whenever I have to work with young kids, I always say, I'm late Gen X into millennial. I always say, look, your generation uses technology. My generation invented technology, which means it represents our way of thinking way more than it represents yours. You hate my music, you hate my clothing, you hate my films, you hate everything about us, but you love our tools. That I will never understand. So what they need to do is recognize that this was us having fun. When we die, all of this should die with us as they create whatever means fun to them. But unfortunately, they're stuck in the world that we shoved them into.
24:13 Does technology erode creativity?
Ryan Vet · There was so much there. You talked about calcification and just kind of this flat line of culture and society. Like the Avatar movie is a great example with that huge gap between, but it doesn't feel like a major breakthrough. The first Avatar was so groundbreaking, and we talked about it forever. One of the other questions I have when it comes to technology in the education space, are we seeing it erode creativity? Are we seeing it impede abilities to be creative thinkers?
Jared Cooney Horvath · I'm not even going to hide that. 100%, yes. So the reason for it, by and large, is most people think higher order thinking skills like creativity and critical thinking and collaboration are separate from our knowledge. That we know stuff and we can do stuff, and we bring those together, off you go. No, it turns out higher order thinking skills emerge from our base level knowledge. They are not separate to them, they literally grow from it, which means the more knowledge you have, the stronger skills you will develop. The weaker knowledge you have, the weaker skills you will develop. And a great example of this is in 2018, the PISA. This is an international exam given to hundreds of thousands of ninth graders across 90 plus countries every three years. It tests them on knowledge, math, reading, and science. Where do you sit on there? In 2018, they also gave kids a test of creativity. It was a creative ideation and innovation test, a standardized test. Now, when it came to the knowledge tests, hands down the top country, Singapore. They have the most rigorous, hardest academic system in the world. They killed it when it came to knowledge. No one was surprised by that. We expect that every year. Great job. Who do you think got number one when it came to creativity? Singapore. And not even close, miles above everyone else. And everyone's like, whoa, how did they do that? Number two knowledge, Japan. Number two creativity, Japan. When we lined up knowledge and creativity, they correlated at 0.92, which meant in a very real sense, the more stuff you knew, the more creative you became. Why? Because creativity emerges from knowledge. So this belief that digital technology somehow absolves us from having to learn stuff, and a lot of schools are using it to avoid, kids, don't memorize facts, just look them up. That's the number one way to make sure you will kill creativity and critical thinking. And then when you bring stuff like AI on board, remember, it becomes outsourcing to most of these kids. So someone ran a study recently where they said, you can use AI for a creative task, something like, here's a brick, how many ways can you use it? That type of task, right? Kids who used AI came up with way more answers than kids who didn't use AI. So people said, hey, it boosted their creativity. The researchers then said, no, the answers they came up with were all almost identical. So it narrowed their creativity. Even though they had more stuff, they said the exact same thing as the group over there. So if we would call creativity originality, there was no originality there. The kids who didn't use AI didn't come up with nearly as many things, but everything they came up with was unique and meaningful and useful in that context. So, again, it's just another one of those examples of, we have to give ourselves time to think. You've got to give yourself time to fight, get bored, struggle, come up with bad ideas. If you look at the history of creativity, right, I'm working on a documentary right now on genius, so this is all fresh in my head. The people who come up with the best work are the exact same people who are coming up with the most work. It's just that the majority of what they're doing is average or okay, it's fine, but it's not great. But greatness comes from goodness. And the more you chug and work and do crappy things, the better chance you have of figuring out what's going to be good. But when you eliminate all of that and just say output is all that matters, a nice, clean, pristine piece of writing, that's all that matters. No, that was never what mattered, especially not when you're a student. Maybe if you're an employee and you're going to get fired if you don't have something, then you think about it. But if you're a kid whose job is to learn, mess up. Try something, suck for a bit. That's okay. That's how you're going to learn to not suck in the future.
28:33 Embrace the suck, and school as a performance institution
Ryan Vet · That's so good. We just had someone on the show, and he's a Gen X and a Gen X influencer, and he talks about all the Gen X stereotypes. And his closing line to everyone was embrace the suck. And his argument was simply that we have removed so much of this learning, so much of the failure. We've made it bad to fail. And I think that parents are swooping in, especially you have this helicopter generation of parents that are swooping in to protect their kids and bubble wrap them from the hard things that they had to deal with, but it's to the children's detriment.
Jared Cooney Horvath · Yeah, yeah. And historically, education was understood as a learning institution with performative features. You're there to learn. But of course, there's going to be an end-of-year test, or there's going to be some summative exam to get you into college. But that's like 1% of your time. What are you doing with the other 99% of it? You're learning, you're screwing up, you're getting a D, you're getting a C, you're getting a B. I'm getting better. Hooray. Somehow, though, I think we've started to interpret school as a performance institution full stop, where you are not allowed to ever get anything under an A, otherwise you're not going to Harvard or wherever the heck else you want to go. And that is 100% where now parents will make sure their kids never fail. I used to teach at a university, I won't name the university. I stopped teaching there shortly after they told us we are no longer allowed to fail students. So long as they pay tuition, they get their degree, regardless of whether they even show up to class. And remember, I worked in a medical school. Do you think that's a smart move, to pump out the next generation of doctors who, so long as they can afford tuition, I'm sending them into the world to do heart surgery? You think that's okay, that's what you want to do? But that's exactly it. We're so afraid of hurting some people's feelings and so afraid of the suck, and our kids may be emotionally suffering from that, that we make sure they never fail. But in so doing, they now have resilience for nothing. They have grit for nothing, and they don't really have the skills we need them to have to actually contribute meaningfully beyond what we're already doing right now.
30:40 Is it too late for Gen Z?
Ryan Vet · That's powerful. And I want to ask a hard, hard question. Is it too late for Gen Z? We see that Gen Z is now as young as about 14, but they're also in our workforce right now. Is it too late for them?
Jared Cooney Horvath · Heck no. This is what I love about the brain. The brain is malleable, it's always changing. You just have to put in the work. And that's what I'm hoping, is that as they get older, they get mad at us and they start putting in the work. They're like, you know what? I'm going to start reading the classics. Why? So I can push back against you. I'm going to start digging into these deeper topics. Why? So I have something to push back with. So by no means. No, that's what makes it good. As long as you're alive and breathing, this whole system is still malleable. So keep pushing it. You want to go in a different direction, the show is always on. So just start pushing the other direction.
31:24 Gen Alpha, Gen Beta, and 31 states pushing back
Ryan Vet · That provides a lot of hope. Now, looking at Gen Alpha and Gen Beta, so Gen Alpha, you know, zero to 13 right now, and Gen Beta, we haven't seen school systems change. Schools are still doing what they've been doing. We're seeing some states start to ban cell phones in schools, while some schools are still giving out Chromebooks or the device of the day to their students. What do you see as the future for Gen Alpha right now? They're mostly in elementary school and starting middle school. Some will be starting high school this fall. What do you see as their future? And what do educators, policymakers, and even parents need to do to equip this next generation?
Jared Cooney Horvath · Well, we're lucky. When I first published the first round of this book seven months ago, there were three states proposing ed tech bills where they were going to start to limit the use of tech in schools, require evidence of efficacy, those kinds of things. Seven months later, there are 31 states at last count that have proposed bills. So it's going really fast, that people are realizing we've got to divest from this ed tech stuff and get back to analog learning. So I think you're going to find probably the most harmed group is going to be late Gen Z, early Gen Alpha. Those are going to be the ones where we're going to have to push hard to get over the addiction, to get over the shallow thinking, to get over the shallow learning, to get over the lack of identity formation. But late Gen Alpha, that's where I think they're going to start growing up again in an analog world. And I think we're going in the right direction. That's great. That's all we need. And that's what I love. Take this all the way back. So ed tech, you could say, started in 1962. Heck, you could say it started in 1922. The very first digital tutor was built in 1922, believe it or not. So we've been doing this for a hundred years, but realistically, 2010 was kind of the decussation point where it went from peripheral in school to primary in a lot of schools. Now tech is the central focus. And it's funny, right around 2010, the big argument is why? Like, why did schools start to go tech heavy there? I keep drawing it back to TED Talks. That was really when TED Talks started to become hyper popular. And I don't know if you remember when TED Talks started, that was it, man. As an academic, I could say the greatest thing in the world, six people will hear me. If you get on that little red dot, you could say the weirdest, silliest, and most nonsensical thing in the world, and millions, tens of millions of people would listen to you and then take that to heart and try and change the world accordingly. If you go back to all the early ones, and in fact keep it going to today, every TED Talk basically has the underlying message: education is broken. Schools are killing our kids' creativity. Kids are not ready to learn, school is broken. And of course, when there's something broken, you have to fix it. And so the suggestion they always gave was technology is the fix. The T in TED is technology. Don't be surprised when they say, here's your cure. Now the problem is, in 2010, when they started this narrative that ed is broken and tech is going to fix it, that was literally the best education had ever been in history. Not only was it not broken, it had never been better. More kids had access to it, more kids were succeeding through it, test scores for what it's worth were higher than they'd ever been, well-being scores were higher than they'd ever been, confidence rates, everything was good. But of course they come along, and it's hard to sell a product without a problem. So they started this narrative that it is broken, and that has been driving everything for 15 years, that we need a revolution. School is broken. But I think we're lucky now that we realize that tech was unnecessary and actually harmed what we were trying to achieve. All we have to do is go backwards. Normally, when you talk about change, you talk about needing to lay down a new infrastructure. You've got to build a system upon which this change can live. We're not talking about change, we're talking about reversion. We're talking about going back to what we were doing 15 years ago. Good news, those train tracks still exist. That infrastructure is still there. It's a little dusty, sure, we've got to dust it off, get that printer out of the closet, but we were already doing it. Handwritten notes, notepaper, blue books, talking, having debates with peers. That's all we have to do. So I think we're in a very lucky position that we can make this change really fast, and then late Gen Alpha into Gen Beta, they're going to experience the good stuff again. Hopefully they'll bounce back up to where we were and then start to surpass where we were in 2010.
35:59 Teacher-facing tech is fine, student-facing is where it breaks
Ryan Vet · I think that's great. You said blue books. I had forgotten all about those, so thanks for that nostalgic trigger there. So, in your professional opinion, then, how far back do we go? Do we go back to the blackboards and chalkboards, or can we stick with the whiteboards still?
Jared Cooney Horvath · Oh, yeah, whiteboards, they do the same thing. There's no difference there. Smart boards, that's where things started to go. You don't need a smart board. You can use it. So a good rule of thumb: if it's teacher-driven or teacher-facing technology, we're pretty much in the clear. So when people say I use PowerPoints to teach, well, cool, you're the teacher driving that. That's fine. It's any student-facing, especially student-facing internet-connected digital device, that's where things are going to go haywire and learning is going to start to collapse. So this is just like, look, we're not saying ban all tech. We are saying get rid of one-to-one computers. No kid needs a laptop, especially not in primary years. But you still have a computer lab. So when it comes time to learn about a program, yeah, we've got a place to do it. But that computer lab is locked down, heavily controlled, largely guided by the teacher. It's just not a free-for-all. And I always say that that's no different than anything else we do in school, right? If you want to learn to play the oboe, you go to the music room. No one says you need to carry an oboe around every class and play it. We understand, when you want to learn the tool, you go learn the tool. You want to learn basketball, there's the gym. You do not play basketball in every class, you play it in gym class. It's the same with tech. I have no problem saying we want to teach kids computer skills. That's a curriculum question. There's a computer lab for that. Sweet. The problem is that somehow that argument has morphed into we need to teach all skills through a computer, otherwise kids won't be ready. That's a pedagogical argument, and that's a very different ballgame, and that's very wrong. Saying we need to teach coding is one thing. Saying we need to teach reading through a screen is a very different thing. One is okay, one, as you can probably guess, has led to our downfall and detriment in education right now.
37:57 Does AI belong in school?
Ryan Vet · That's a really important distinction. And so as we kind of wrap up here, I would love to know, AI is not going anywhere. That's our new world. That's where we're going to be. A lot of people are saying, hey, we need to teach prompting, we need to teach logic, we need to teach X, Y, and Z. What is your perspective on AI in schools, and where it has a fit and where we need to be cautious and slow down?
Jared Cooney Horvath · It fits absolutely nowhere in school. AI is not a learning tool. It never has been. You can't pretend it is. And in the argument that it's here, it's not going anywhere, well, cool, fentanyl is here. Does that mean we need to start giving kids fentanyl in school so they figure it out? No, we realize there are some things that exist in the adult world that kids are not ready for. There's very rarely a time that I'll use the B word, ban, but I will use it there. Smartphones, don't need them. AI, don't need it. Why would we ban that? Because think about it, school is nothing other than a giant ban. It is a recognition that kids are not ready for the real world. So what we are going to do is expose them to ever more increasingly complex ideas until they are ready for the real world. And what that means is we ban everything. We ban clothing, we ban language, we ban actions, we ban movements. You can't go to the bathroom unless you ask for a pass. We ban what you are allowed to say, we ban what clubs you are allowed to join, what you're allowed to learn, what you're not. Everything is a ban because we are cultivating experiences to make a healthy adult. AI has no place in any of those experiences until you are an adult. And to say that we need to teach kids prompting, what the heck are you talking about? Prompting? You mean you need to teach kids to ask a question? I have a three-year-old, pretty sure she's got that just fine. Now, you made a good point. What do we need to teach them? Logic? Sweet. You do not need technology to learn logic. And in fact, when you try and learn logic using tech, your logic skills are worse than people who do it analog. You made the great point. In order for AI to be useful, you need to bring stuff to it. What are we bringing to AI to make it a meaningful tool? That is a K through 12 general education. That is why we don't use AI. Imagine I invented a machine that lifted weights for you. Now, as an expert weightlifter, that might be just fine. That's all I needed, was to show you that this machine could do it. But if you're someone trying to actually get fit, the machine is completely not fit for purpose. That's what AI is. If your goal is to learn and develop and understand the world and be able to think and maneuver your way through it, that is just not a tool for you. So I see there being no place for AI in school one way or the other. And I think we're lucky enough, I think that whole bubble will pop sooner rather than later. So we won't have to worry too much about it much longer.
40:50 Recultivating an environment for learning
Ryan Vet · Well, I hope you're right about that, Jared, because having kids in school and going through school, I want them to learn. I want them to learn that human-to-human interaction. I want them to learn and have the struggle, and maybe even bring the blackboards back, so if they get in trouble, they have to clap out the chalk erasers in the back. Not that I ever did that in the grade, but this is, I think, a good point.
Jared Cooney Horvath · That's where you just see, we have to recultivate an environment for learning. That's it. If you think of it, a hospital has a very clear environment, which causes you to behave in a very certain way when you're in a hospital. You know you're in a hospital no matter where you're at in the world. Libraries have a very clear context which forces you to behave in a certain way. We just need to bring that back to schools. And if that means chalkboard and chalk is the way to make sure we're now cultivating the behaviors we need for learning and development, by all means, bring it back.
41:43 Where to find Jared and the book
Ryan Vet · Love it. I love it. Well, Jared, your book is coming out, so everyone needs to order that. Order it now, order it this week while you're listening to it, or get on the pre-order if it's not quite yet released. How can people find your book, and how can they find out more about you and your Substack and everything else?
Jared Cooney Horvath · Yeah, so the book is called The Digital Delusion. And if you just look it up, it'll be on Amazon and all the bookstores. It's in a lot of places. And the Substack is also called The Digital Delusion. So that's on there, we just write new things. Here's new data emerging, here's new research coming out of this lab. I just try and keep you up to date on what's going on. And I also have a YouTube as well, called Jared Cooney Horvath, if you look that up. And we just put out learning videos. Here's new research, here's new ideas. So just trying to kind of keep the story going. Now that the book is done, just kind of keep people updated on where things are standing, who's doing what right now.
42:31 Close
Ryan Vet · I love that. Well, Jared, thank you so much for your insight and for just the impact you have and the conversations you're having about learning. I think this is really important as we talk about the next generation, the future of our culture and our society. We've got to invest and pay attention to what they're learning and what might be distracting them from core learning. So, Jared, thank you so much.
Jared Cooney Horvath · Thank you.
Ryan Vet · And thank you for diving into this topic.
Jared Cooney Horvath · Absolutely.
Ryan Vet · And thank you to all of our listeners. Be sure to like and comment on this episode, and we look forward to seeing you next time.
Ryan Vet · Thanks for tuning in to the Ryan Vet Show. Be sure to subscribe, comment, and like this episode. Plus, share it with someone who needs to hear it.
Full show notes
Cognition had gone up every single generation since the late 1800s. Literacy, numeracy, working memory, executive function, IQ. School was the engine, and each generation spent more time in it than the one before. Gen Z spends more time in school than we did and is the first generation on record to move backward. Dr. Jared Cooney Horvath’s argument is that something decoupled cognitive development from schooling, and that the something was the technology we put in front of every child.
Horvath is a teacher turned neuroscientist who directs LME Global and testified before the US Senate in January 2026 that pulling smartphones out of schools is only step one. His book The Digital Delusion, re-released by Harmony on August 18, 2026 with roughly 25% more material, lays out the data and then explains the mechanism: transfer. Learning only travels hard to easy. You learn to handwrite before you type and to do arithmetic before you reach for a calculator. Technology exists to remove friction, so a child who starts easy has nowhere harder to go.
That leads to the distinction at the center of the conversation. An expert who hands a task to a tool is offloading, because their expertise lets them vet the output the moment it appears. A novice doing the same thing is outsourcing, and learns no facts and no skills at all. His illustration is a rented e-bike that runs out of battery: the fit rider pedals home, the unfit rider calls an Uber.
Host Ryan Vet, who does not allow screens in his own house, presses on what that means for parents and schools. Horvath’s answer on AI in K through 12 is unambiguous. His answer on whether Gen Z can recover is two words.
In this episode:
- Why Gen Z is the first generation on record to score lower than its parents, and what decoupled cognition from school
- Transfer, and why learning only runs hard to easy
- The difference between offloading and outsourcing, and the e-bike that explains it
- Why AI tutors mostly make kids better at AI tutors
- Physiological synchrony, the 0.68 effect size, and what a screen has no way to replicate
- Why millennials are more digitally literate than Gen Z despite never being taught
- What PISA’s first creative thinking results say about knowledge and originality
- The line between teacher-facing technology and student-facing devices
- Why Horvath says AI belongs nowhere in K through 12
- Whether it is too late for Gen Z, and why the fix is a reversion rather than a revolution
