The False Promise of Tech and AI in Schools | Jared Cooney Horvath, PhD, MEd

[Kris Perry]: Welcome to Screen Deep, where we decode young brains and behavior in a digital world. I’m your host, Kris Perry, Executive Director of Children and Screens. Today’s guest has become one of the most prominent voices challenging many of our assumptions about educational technology, Dr. Jared Cooney Horvath. He is a cognitive neuroscientist, former classroom teacher, director of LME Global, and the Director of the Science of Learning group. His work focuses on translating research on the brain and learning into practical applications for education. His recent congressional testimony on ed tech and children’s cognitive development sparked widespread national discussion. He’s also the author of The Digital Delusion: How Classroom Technology Harms Our Kids’ Learning and How to Help Them Thrive Again. Today we’ll explore what the evidence tells us about technology in the classroom, why learning and productivity aren’t always the same thing, what AI may mean for the future of education, and how parents and educators can help children develop the knowledge and skills they’ll need to thrive.

Let’s go right to the heart of things, Jared. Your testimony before Congress earlier this year went viral for powerfully making the same point contained in your book, The Digital Delusion, namely that the promise of using technology for learning is hollow and that schools have mistaken technology adoption for progress in education. What’s the evidence that this mistake or delusion is real?

[Dr. Jared Cooney Horvath]: I think – and you’re diving right into it – it’s gonna start with correlative evidence, where when we look at standardized tests – international, national, doesn’t matter. Math, science, reading, doesn’t matter – we started to see a pretty significant downslope right around 2010, 2012, and it just kept getting worse and worse. COVID hits, it got super bad. And everyone thought, “Oh, COVID’s done, it’ll come back.” Nope, still going down. So of course, we gotta dive in and say, “What the heck is changing? Like, what’s actually going on in schools that’s kind of mirroring this?” And it can’t be schools themselves. We’re told every day how school looks the same now as it did a hundred years ago. Cool, then don’t blame school. It can’t be evolutionary biology. We haven’t – 10 years isn’t enough for any of this to change. What the big kind of takeaway seemed to be was it’s the tools we were starting to introduce into school that really started to push things down.

So in basically every data set you see with international testing, when you correlate tech use in school for learning purposes with performance on these exams, learning starts to go down. So once you get all this correlative data pointing in one direction, you’ve got to say, “Cool, let’s go to the research and see does the academic work back this up?” And what you see is since 1962 we’ve had data on digital technology and human learning, and it all points in the exact same direction. It’s a weaker form of learning than basically every analog technique we’ve already been doing. And from there, then, I always say the third step, if you really want to make a good argument, is mechanism. Is, you’ve got your correlation, you’ve got your academic research, now you need an explanation that explains all of this better than any other explanation out there. And luckily that’s where my expertise in biology and learning come in, where I can say it’s because human beings learn like this, do you now see why that would likely drive the data in these directions? More so than any other excuse people are given.

So I think the data’s clear. We got correlation, we got academic, and we got mechanism all point in the same direction. It’s a very hard story to overturn right now.

[Kris Perry]: So I’m also thinking about how teens are on screens for around eight hours a day and they’re in and out of school. How hard is it to disentangle the effects from screens and schools from the impacts of screen use more broadly?

[Dr. Jared Cooney Horvath]: This is where good data and good research will really come in. You can assume on a big enough database, all the other things will wash out. Like, let’s say we have a data set with a million students and the only big difference between them is the amount of time they use tech in schools for learning. You can run regression analyses and stuff, but what you’ll find is with that huge of a data set, everyone is already using tech outside pretty significantly. And all of them are already on social media pretty significantly. And you can run those correlations and see it doesn’t correlate with learning. Thank goodness we have such big data sets with international data that you can really disentangle what seems to be one of the driving causative mechanisms here.

And to be fair, it’s a multifaceted problem. As you said, it’s family life, it’s social life, it’s emotional life, it’s what tools you’re using out of the home, it’s nuclear family, all of these things are going to feed into it. But I think at the end of the day, from the pure cognitive aspect, this is a big dog. This is one of the major bricks in the wall that we can point to and say, “Here is something we can address tomorrow.” So let’s focus there.

[Kris Perry]: I love this quote from your book, the claim that “schools should teach computer skills” has morphed into “schools should teach all skills through computers.” What important components of teaching and learning do you feel are most displaced by digital tools?

[Dr. Jared Cooney Horvath]: The – hands down, the most displaced is what we call “affective empathy,” or biological synchrony. Human beings have evolved 150,000 years to learn from each other. That’s one of our superpowers. We have built-in biological mechanisms that says when another human being is doing something, you can resonate with them and learn from them. You don’t even have to do it. Just simply by being here, you will learn from them.

So, in a very real sense, if you and I started to synchronize right now, our hearts start to beat simultaneously. We start to breathe at the same rate. We blink simultaneously. If I could image our brains, our brains start to look almost identical. You can now understand that’s synchrony. If I’m a student learning from a teacher, in that moment, I’m no longer learning from the teacher. I’m thinking like the teacher. I’m in that teacher’s head. We’re making the same moves together. So one of the biggest drivers of learning we have is the human connection. 

Now, surprise, surprise, it is real hard to get that same resonance on a screen, and even harder when there is no other human on the screen, when you’re just using a tool for tutoring or math or language. If there is no other pair of biology, there can be no resonance. An entire aspect of human learning just basically gets dumped.

So I’d say the biggest displacement is we are built to learn from each other. That is what makes humans unique. Remove the other from the equation and learning becomes little more than a data dump. And in which case, tech is still not the best way to do a data dump. There’s still better ways to do it than tech.

[Kris Perry]: You know, you’re reminding me of how important just the dyadic relationships are in our lives, starting from the first days of life and going all the way through middle school and high school. That synchrony is so key to our humanity and what differentiates us from other species is that deep connection and how it translates into greater learning and better outcomes.

So one thing about learning is it’s not the same thing as productivity. They’re different. How does the ed tech marketing use that distinction? And how are they confusing us when it comes to those two ideas?

[Dr. Jared Cooney Horvath]: Yeah, a good rule of thumb is that the tools an expert uses to produce work are very different than the tools a novice needs to learn how to become an expert. One is a developmental process, one is a output-driven field or endeavor.

Digital technology, if it can be said to be anything, is there to make production easier. It makes reading a map faster. It makes calculating numbers quicker. It makes word processing smoother. Now that’s all fine if you’ve already developed expertise and you’re just pumping out emails, you’re pumping out data, you’re pumping out whatever. But when you’re a kid who doesn’t know how to write yet, using a tool that makes writing easier is counterproductive to what you’re trying to achieve. If I’m a bodybuilder and I decide today to use a machine that’s gonna lift weights for me because I have an injury, sweet. If I’m a kid learning how to lift weights and I use the same machine, that’s the best way to make sure I’ll never actually develop that muscle. So I think, yeah, when we remember that tech was never developed for learning purposes. Nothing about tech is designed for learning and development. It’s designed for offloading and production. 

And go back to the very first one-to-one movement in schools. The first Chromebook that was sent into schools, it was done – that Chromebook was attempting to be sold to business people. It was just a very small laptop they wanted business people to use. Turns out business people didn’t like it. So what did they do to recoup costs? They said, “Let’s sell it to schools and give it to kids.” It wasn’t designed for them. It had nothing to do with them. It was just a way to make sure they didn’t lose a ton of money on this product that nobody really wanted. So even when you go back to the roots of this, you saw it was never a learning endeavor.

And so that’s the argument that they’re gonna make, is because adults use computers to do work, your kids need to use them in school. Mate, no one would make that mistake anywhere else. Like, think about it. As an adult, I do not use floaties when I’m in a pool. Does that mean I shouldn’t give my three-year-old daughter floaties when she’s in a pool? Because as adults, that’s not what we do. Adults and kids are two different species. We are trying to live, they are trying to figure out how to live. That’s a different ball game requiring different tools, different thoughts.

[Kris Perry]: Huh. I mean the business model – I hadn’t thought about that layer of it before. I know the business model that wants to keep us engaged and the algorithm gets more and more sophisticated at keeping us there, but I hadn’t thought about even the hardware and the business argument for the hardware.

Earlier you were talking about tech adoption, new tech adoption in general. It’s not just in educational settings, usually it revolves around reducing inefficiency or points of friction in tasks in daily life. But are inefficiency and friction in some ways valuable for learning?

[Dr. Jared Cooney Horvath]: They’re not valuable for learning, they are learning. Two words that have never been linked to human learning: speed and efficiency. Those are not synonymous with learning. By all means, again: output, production. Faster, better. Cool. No one cares what you were thinking. No one even cares if you did it. If AI popped out an article and you’re a journalist and everyone loves that article, I don’t think anyone’s going – I mean, we will flip out at you, but I don’t think your boss is gonna say no. They’re gonna publish the dang thing. 

But I don’t care about output from my kids. We don’t learn to write essays because I genuinely think your essay is gonna be in the newspaper. You’re figuring out how to organize your ideas. I’m not teaching you history because you’re gonna become a historian. I’m teaching you how to think through knowledge.

Now, when we bring speed and efficiency into that process, all the little nuances go away. All the little moments of actual development, the “aha” moments, the confusion, the, “But what if I always thought it was this?” All of that is learning. So by all means, the more you remove friction from learning, the worse learning will become. So it’s not surprising that when we use a tool that’s the simplest, most frictionless tool in the world, kids learn less. It’s just not a learning endeavor.

[Kris Perry]: I mean, it almost makes sense that you would teach kids about learning. So you don’t just teach them content, but you teach them the process of learning. That the process of learning is that you go through periods of struggle, that you go through periods of friction, that it’s messy, that you feel like you’re failing, it’s too hard, you’re stretched, you’re digging deep, all of those analogies that get at why the friction is good for you. And we definitely understand it better when it comes to our physical health but it seems like kids have less knowledge about how it’s good for their cognitive and academic success.

[Dr. Jared Cooney Horvath]: Can I share? And I think you’d like this. A long time ago I was asked to teach a study skills class to kids. So I did. And at the end of that class, like 90% of kids could tell you what to do. “Here’s a good strategy, here’s a bad strategy.” But when you followed them, only about 30% of kids were actually doing it. So everyone knew what they should be doing, no one was doing it. And for a long time we kind of battled, “What’s going on here?”

And the answer was pretty simple. It was exactly what you said. The next year, rather than teaching study strategies, I taught how learning works. How does memory work? How does attention work? What’s surface learning? What’s deep? How do we move? I never even mentioned study strategies. Lo and behold, after that program, 65% of kids were using successful study strategies without me ever saying them. And that’s when we realized learning is a biological process. Like digestion, like physical health, like breathing. But people treat it as something else. They think it’s ethereal. They think it’s unknowable. And some people are better at it than others. The more we can demystify learning and say, “No, no, no, it’s biology. It’s the exact same process as everything else. Here’s what it’s gonna feel like. Here are the steps you’re gonna have to go through,” the more likely kids are to kind of embrace it and take agency over that process. So I think you’re spot on with that. How do we focus back on learning as a process as opposed to just learning as a maybe-you’ll-figure-it-out-if-I-just-teach-you-well-enough.

[Kris Perry]: It almost makes me want to go back to school so I could actually go back to school and understand that as uncomfortable and painful and stressful as that is, it’s good for me. So thanks for clarifying that it’s like digestion.

So speaking of thinking and learning processes, we’ve learned from other Screen Deep and Ask the Experts webinar guests that multitasking is actually not possible cognitively. That what people call multitasking is really very fast task switching. Is classroom tech use affecting multitasking and cognitive attention?

[Dr. Jared Cooney Horvath]: Absolutely, and it can’t not, because you gotta think about it. The exact same tools we use in the classroom are identical to the tools kids are gonna use at their home. And in fact, in 88% of cases, it is the same computer. It’s not just similar laptops, it is the school-issued laptop they’ll be using at home.

Now, when you track how kids use computers typically, every year, about 2,500 hours every single year, they will use the tool to passively consume rapidly switching media content, which is multitasking. Now we try and say, “Can you learn?” Well, when I’ve just spent thousands of hours training myself to do one thing on a tool, and you hand me that tool and say, “I need you to do something totally different,” most people won’t be able to do it. That’s why kids, when they use a computer for studying at home, make it about six minutes before they start multitasking. When they use a computer in a classroom with teachers present 24 to 38 minutes of every hour, they’re off task multitasking. It’s not because the next generation is morally weak or that they lack some cognitive ability. It’s a pure training thing. If you assume it takes 10,000 hours to master anything, then when your kid is four years old they have mastered one thing: multitasking on devices. Now when we bring it into a class and we would say, “Don’t do that here,” they go, “Mm-hmm,” and then go right ahead and do it anyway. 

So yeah, multitasking is the worst thing you can do and these tools – that’s the attention economy, they beg you to multitask. That’s how they make their money. So now you can start to see another that just pure training won’t work. 

I would say that alone should have been the end. To me, that data point alone is, “The battle’s over, the war’s done. Go home, everyone.” The second digital technology became an entertainment tool instead of an educational tool, trying to pretend that we can somehow maneuver this to do something different, it’s going to be a losing battle every time.

[Kris Perry]: Yeah, so we might as well just abolish the word multitasking because it’s also not a thing, right? So this idea that our brains can switch between the two purposes of one device is a really important thing to underline, like you have, because I think we all underestimate the power of the entertainment aspect of the device, especially in an educational setting.

You participated in a research retreat on educational technology with Children and Screens in 2022, and you shared some powerful insights there. But that was before the more recent rapid evolution and widespread adoption of AI. In such a short time, AI has made some seemingly fundamental changes to the ways students are writing, learning, and thinking. I understand you are re-releasing your book with updated content on AI, so I wanna dig into AI specifically. 

[Dr. Jared Cooney Horvath]: Yeah.

[Kris Perry]: Do you see generative AI as an extension or intensification of the same ed tech problems we already talked about or as something qualitatively different?

[Dr. Jared Cooney Horvath]: I would have to sit on extension. It’s just the same story writ larger. So I can’t see it being any different, especially if AI is run on a screen. We’re gonna hit the exact same problems we’ve always hit. The thing that AI adds to the equation is it just makes everything worse.

It just – AI is the silliest thing. I always call it the tool nobody asked for, solving problems nobody had. What do these things exist for, these large language models? The people who invented them flat out said, “We don’t know. We gave it to the world and said, ‘You tell us what it’s good for.’” Everyone appears to have said coding, and now they’re starting to say actually it’s not very good at coding. What do they do? Who cares?

So this is one of those great reasons where if you had to say AI is good for anything, it’s good for expert, what we call, “offloading.” That is very different than novice outsourcing. So here’s the difference. A good kind of example is, my sister-in-law is fit. She’s, like, a gym rat. She’s always working out. She’s one of those people. Her husband is not. That dude is – he walks to the fridge. That’s about as far as he’s gonna make it. They’re a very funny couple. I love them together. Anyway, one day they decided to rent those e-bikes, those electric bikes, like motorcycles, really, to go tour around Melbourne. So they’re going around Melbourne and eventually their e-bikes run out of batteries, like they always do. Now, in her case, it doesn’t matter. Why? Because she had the underlying fitness. When the motor stopped working, she just pedaled home. That’s called “offloading”. When you’re using a tool to do something you can do, you just don’t want to do it in the moment, but you can pick up as soon as you need to, you’re offloading. He, on the other hand, when his motor ran out, he had to call an Uber to take him home. Why? Because he had no ability to actually use the machine without the motor. We call that “outsourcing.” When you use a tool to do something you can’t do, and now when the tool is gone, you got nothing.

AI works, if you can say it works for anything at all, because experts can vet the output immediately, meaning it’s an offloading tool. Like think of me, I’m good at stats. I could do stats, but it’s gonna take me two hours to run a data set here. AI could do it for me in two seconds. Sweet, I’m gonna use AI. But it only works because I can immediately vet the output. When it spits out a number, I can immediately go, “Yep, that’s right,” or, “No what? No, that’s wrong. Did I type in something wrong? Something’s happening.” Give that same tool to a kid and say, “Now use AI to run statistics.” When that spits out a number, they have no knowledge, they have no background, it’s just a number to them. They copy it, they paste it, they won’t have known was it good, was it bad, and they certainly won’t have learned the process of stats in between. So when experts and novices use the same tool, it’s offloading for us, but it’s outsourcing for them. It’s the best way to make my life easier. It’s the best way to make sure those kids will never learn a thing. And now they will forever be tied to the tool. If they ever want to do stats, it’s a matter of, “The tool has to be there, whatever it says is right because I don’t know any better.” It just becomes a copy-paste exercise.

Now tell me, who does that sound good to? Parents, students, or tech developers? Hey, who benefits when kids become dependent upon technology? It’s not a mystery. That is literally what they are trying to do. So when we see it, and now that we see it, we can push back and hopefully do something different.

[Kris Perry]: Offloading and outsourcing are clearly the key issues for AI in education and nobody seems to be directly addressing it before implementing AI tools in schools. And as you point out, they’re selling it as a solution to a problem. What are the risks of overhyping AI for education?

[Dr. Jared Cooney Horvath]: Turn on LinkedIn, man. It’s – it’s everywhere that if you’re not using it, you’re a bad teacher. That if your kids aren’t using it, they’re – but of course when your kids are using it, they’re not using it to cheat. Of course they are. Dude, I do not know how much more data we need on this. We now have four years of data showing that when kids use AI, they cheat. Guess what? I have four years of data showing that when kids get a cigarette, they smoke it. We can’t be surprised when that happens. That’s literally what the tool is built for. 

A lot of people come out and say, “Oh, you could use it to do this. Like, kids could use it to ask them probing questions.” Yeah, mate, I could use my car to improve my fitness by popping it into neutral and pushing it to work every morning. Of course, I’m never going to do that because that’s not actually what the tool demands of me. That’s not what the infrastructure is saying. It’s the same with AI. It doesn’t matter how you could use it, it’s how are people actually going to use it. How is it designed to be used? And it is not designed to be used to bounce ideas and learn off of. It’s designed to outsource for them. So they’re gonna use it to cheat. That’s what they do. 

Now, what does that mean, kind of writ large? Why we never stopped – now, and to be fair, I think AI was big tech’s –  I think that’s when they shot themselves in a foot across the board. Because up until then, schools had never really stopped to ask, “Why?” Here’s a new tech, give it to kids. “Okay.” Here’s a new program that’s gonna help math, give it to kids. “Okay.” Here’s a new program that’s gonna help languages, give it to kids. “Okay.” We never stopped to say, “Does it work? Is it safe? Is it legal?” We just said, “Give it to kids.” AI came so fast and so furious, it was the first time that a lot of schools stopped and said, “What are we doing? Why?” Like, do – it’s not a question of, “Do I give this to my kids?” It’s a question now of, “Why would we ever do any of this? What’s the evidence? What’s the use cases?” And I think what’s gonna happen is that wake up call is gonna lead a lot of schools to start pushing back on all tech. Because AI opened the door to them to say, “Wait a second, we have no data for any of this. What is going on here?” Now go bigger. I think AI is going to be the thing that pops big tech. Every thing comes and goes.

So I’m doing a documentary right now on genius, and one of the funniest things is to watch genius cycles. If a genius cycle pops up, it will disappear. It has to. The latest genius cycle was Silicon Valley, about 25, 30 years ago. It has to die. I think AI is its death now because it was the first time everyone said, “What are you doing?” And now they’ve invested so much money to try and hyperscale this tool that China comes in and says, “We’ll do it for a sixth of your price, and they can own it.” Fine, here you go. So I think this was their Icarus-flying-too-close-to-the-sun. They’re gonna harm their position in education, they’re gonna harm their position in the workplace, they’re gonna harm their position worldwide. They just gave us all the reason we need to stop and say, “What are we doing?”

And so there you go. So I think not only are they shooting their foot in school, I think this is a bigger story than that.

[Kris Perry]: If you were a classroom teacher, would you ever assign an essay again? A take-home essay?

[Dr. Jared Cooney Horvath]:  Yes, I would. Because here’s the thing most people don’t consider: why is AI basically useless in K-12? Because I know my students. AI is really dangerous in university where I have six hours total with 500 kids. I don’t know their names, I don’t know what they’re into. When they hand in AI, looks good to me. When I’m spending an hour a day, every day with the same kids, I know what you think. That’s what people don’t realize with teachers. We don’t assign things because we genuinely think you’re going to blow our mind. We assign things because we know exactly what you’re going to give us. And that is exactly what you need to give us so that I can bring you to the next stage. And I can predict that, too. I honestly say in my entire teaching career, I’ve been surprised maybe three times by what a kid has done. The rest is highly predictable and guided by me, they just don’t know I’m doing it.

So when we assign an essay to a kid, check this out. Never in my entire K-12 career did I ever say, “Write an essay, hand it in.” I do that at university. I never did in K-12. What we did in K-12 is, “Let’s focus on just the introduction. Today let’s work on just the first paragraph. I want you to go home, practice it, come back with a concluding sentence of your paragraph. Now let’s take it, let’s fix it, let’s cycle it, go do it again.” It’s not just, “Gimme an essay,” it’s a step-by-step thing. I’ve been following you the whole way. If magically you come back with an assignment, an essay that’s really good, but very different than what you’ve been chugging away at for the last month, I’ll know that immediately. I’ll say, “Dude, this is not you. Stop trying to pretend using AI.” It’s the process.

So I always say, because we know our kids in K-12, it is much harder for AI to actually interface between us. University, that’s where you’re gonna hit some problems. And that’s where I know I would not assign a take-home essay a lot of times at university level.

[Kris Perry]: Again, this is an example of how communicating about how we learn while teaching the content – both the process and the content for the student of any age – is probably gonna be more and more critical to them wanting to do what you want them to do. Because they will see it as part of a longer trajectory of effort that they’re scaling up for and will eventually land them maybe in a college classroom or in the kind of job they want or in the kind of relationship they want.

[Dr. Jared Cooney Horvath]: You have to learn – they have to see that. It’s like, the number one question kids will ask is, “Why do I need to learn this?” And every teacher will be confronted with that. And a lot of the times we try and invent some weird future where it’s like, “Imagine if the satellites collapse, you need to be able to…” and I always say, “No, no, no, no, no. The reason you need to learn this is because all learning is additive. If you don’t learn this, you can’t learn the next thing. And if you don’t learn the next thing, you can’t learn the thing after that. And eventually you keep going up that ladder, that’s when you’re gonna learn the thing you literally want to do with your life. But if you skip this, you cannot jump to that. You can fake it, you can pretend – that’s like giving everyone a phone with a camera and now everyone thinks they’re a photographer. Nope. There’s a lot you gotta go to to get to the stage where you’re actually doing the work. This is that work.” So I’m always like the king of when people say, “Why do I gotta learn this?” I say, “What do you want to do with your life?” You know, “I want to be an engineer, I want to go to the moon.” “Sweet. You can’t unless you learn this.” “But I will never use it.” “No, but you can’t learn the thing you will use unless you have this on lockdown.” That’s how learning works. It’s all constructive.

[Kris Perry]: One thing you do in your book that I find really helpful is to respond to the common arguments that people, not just ed tech companies, but also parents and educators, make to advocate for tech use in schools. These include: “Tech is the future, and the kids or school would be left behind if they don’t use it,” or, “It’ll work great once the teachers get more used to it,” or the generic, “The genie’s out of the bottle, digital devices aren’t going anywhere, so we may as well cope.” What are your responses to these arguments?

[Dr. Jared Cooney Horvath]: I think man, we could hit each one in turn, but I think the biggest question – like, a lot of them will fall under two umbrellas. One is the promissory argument: “Tech could be good, you’re just using it wrong.” And that’s the easy stuff where you go, “Look, when 90% of people are using it 90% of the way in this way, that can’t be wrong.” What you’re really saying is, “We’re not using it the way you wish we would have used it.” Cool, man. Your intentions are irrelevant. The guy who invented dynamite, Alfred Nobel, intended his dynamite never to be used in war. He said it flat out, “No one is stupid enough to do that.” What’s the first thing we did? So do you think his intentions were more important than how we actually used it? The guy who invented the television said this is meant to be only in classrooms, it’s gonna bring knowledge to the world. What happened? It went to the home and it became a tool for advertising. Do you think his intentions overwrote what everyone did? No, so intentions are irrelevant here. It’s what are people actually doing? And that’s what we’ve gotta now address.

But to the other side, the, “How will our kids ever get a job?”-based arguments. The one thing that I love more than anything is that Gen Z is less digitally literate than millennials. Gen Z was raised on tech. We were not. We had a tech lab. This is where you must realize digital literacy is very different than digital use. By all means, our kids know how to use these machines, sometimes better than us. But use is irrelevant. Orson Wells said, “I could teach you how to make a film in a day and a half, but it’s gonna take you decades before you have anything meaningful to say.” That’s literacy. That’s what we’re talking about when it comes to tech literacy. It’s not the tool itself, it’s what are you bringing to the tool? What do you have to say with it? What questions do you understand? What answers are you seeking?  All of that does not come from training on the tool. All of that comes from a general K-12 education. Learn how to think. 

So a good kind of rule of thumb: when we teach kids how to use a tool, they learn how to use that tool. But when we go back to focusing on teaching our kids how to think and how to learn, they now can adapt to any tool. Why? Because I know how to think and learn. So that’s why millennials, you didn’t have to teach us it. They taught us typing. Now I’m better at it than other kids. Why? Because they also taught me how to think and I have a lot to say. It’s not about training them to speak more and more. It’s about what do you now have to say? And all of that does not require tool training. In fact, the more we focus on tool training, the less time we have dedicated to actually building that knowledge that will give them the questions, the answers, the way of thinking that will make them meaningful in the workforce.

So if you genuinely want your kid to be future ready when it comes to work, do not train them on today’s tool. Do not try and predict tomorrow’s tool, because you will be wrong. Remember, no one saw AI coming four years ago. Training kids on today’s tools is like teaching kids in the ‘90s how to rent from Blockbuster. Seemed like a smart idea, totally useless two years later. So don’t worry about tool training. Worry about thinking and learning, and they can adapt to any tool just like we can. That’s what makes us powerful.

[Kris Perry]: One thing you said in 2022 at our research retreat that’s really stuck with me is that there’s no such thing as a learning style, and that often people point to different learning styles to justify the use of new technology. Can you explain that remark?

[Dr. Jared Cooney Horvath]: So, if we go down to the very base, every human being learns in exactly the same way. And that comment makes most people kind of jar, but bear with me. Every human being digests in exactly the same way. Of course they do. That’s self-evident. No one has six stomachs and chews cud, no one spits acid and absorbs nutrients through their skin. We all have the exact same process. Now, by all means, what we put through that system might change. Some people can do gluten, some can’t. The supports you need might change. Some people might need a feeding tube, some don’t. But the process by which the body extracts nutrients and excretes waste is exactly the same for everyone. If you did it differently, you wouldn’t be human. That’s the definition of human biology. All human beings breathe in exactly the same way. Of course we do. Some people breathe different things – they vape, they smoke. Some people need an iron lung. We need different supports, but the process by which we take in oxygen, exhale CO2, is exactly the same for everyone. All human beings learn in exactly the same way. It’s very common to think that learning is some special behavioral personality-based skill. It’s not, it’s biology. And if you are human, the process by which you take in, embody, quantify, combine, utilize knowledge, check that knowledge, is exactly the same for everyone.

So once we can kind of get back onto the base level of learning-is-a-process, and if we know that process, we can make better decisions. Then you can start to say, “Well, okay, maybe learning styles comes at the input. Just like some people can eat gluten, some can’t, maybe that’s where learning styles is.” But it turns out the research says no. By all means, there are going to be some limits. If you are blind, I cannot show you a picture. So there will be some hard limits, but when you take those limits away, it appears when we align instruction to the content rather than to the learners, everyone learns better. If I want to teach you art history and I do it by showing everyone pictures, everyone learns better, regardless of if you think you’re an auditory learner or a kinesthetic learner. And this is where, now, we start to say, “There is no learning style, there’s learning preferences.” People choose, people like certain things. People believe they will learn better certain ways. But the data keeps coming back and saying, “No, when we align to what you prefer, you actually learn less than when we align to what the content demands. Even if you don’t like it, you will learn more going that other route.”

So this is where now you see – bring this back to ed tech. Ed tech, one of its big selling points is it’s adaptive to the kid. “We will give it to you in the way you want it, however you want it. You can choose your own path.” All of that seems good, but it ain’t great for learning. And that’s where we see a structured pedagogy, a structured curriculum done by people who understand the process will always come out better than a free-for-all, choose-your-own adventure that a lot of tech is kind of feeding into.

So yeah, learning styles don’t exist at the biological level. They don’t exist at the mental level unless we consider them preferences, but then preferences are largely irrelevant to what is good learning.

[Kris Perry]: The theme, the pattern of this conversation is that struggle is so baked into learning. And the more that the struggle is relieved or postponed or done by a machine, the less likely it is you’re going to learn. And I think we’ve all been operating from this assumption that learning is very unique to us and we have to have it done in a certain way and these tools are gonna make us better learners and it’s really helpful to have you unpack how much more we have in common, how similar we are, that we don’t need the tools to do what it is we need to do, which is struggle and find the the tools within ourselves to digest the material that’s given to us in school or at home.

[Dr. Jared Cooney Horvath]: Do it together, too. As a lot of research is coming out, your listeners have no doubt heard the concept of “differentiation,” where every kid needs their own style, their own teaching and this way for each kid. And it turns out, no – class-wide teaching works a treat. Even if you are up here and I’m down here, when we’re going through the same process, we start to resonate with each other and learn from each other in those moments, as well.

So, I mean, unless you have huge discrepancies where you have, like, an eighth-grader level math versus a first-grader level math where you might just be, “I need completely different content here,” for the most part, a lot of this idea of individualized learning – it’s not stacking up. Social learning seems to be a lot stronger for us.

So, and that’s – again, that’s what ed tech’s biggest thing is: everyone can just go home and learn on their own. Well, if that was gonna work, surely YouTube would have broken education twenty years ago. It didn’t. Why? Because there’s more to the picture.

[Kris Perry]: Given the historical data about tech in schools and cognitive development, is your position that schools optimally would have no screens or tech used for teaching?

[Dr. Jared Cooney Horvath]: No, not at all. Believe it or not, I’m not anti-tech. I’m just whatever’s gonna work best for learning. 

There are two times I will use the hard B, “ban.” Smartphones – I don’t think they have any spot in school. No one’s ever made an argument why they need to exist – and AI. I have yet to hear an argument why that needs to exist in school.

But beyond there, it’s not about banning, it’s about control. This is where I say – maybe primary years I could get away without doing any screens. I don’t know why a second grader would ever need a screen in his or her face. But when we get up into middle and high school, cool, we have a computer lab again. And this is how we differentiate between curriculum and pedagogy. If we believe we need to teach kids computer skills, say typing, say how to use Excel, maybe how to prompt engineer if that is meaningful to anyone. Cool, we have a lab where we can do that. It’s not your computer. It doesn’t force everyone to have a computer now in front of their face. It’s the same with, “You wanna practice the oboe? There’s the music room.” “You want to learn basketball? There’s the gym.” “You wanna learn computer? There’s a computer lab.” By having it set off and structured, this now demands everyone use actual judgment when they choose to bring tech in. So the tech is still there, and a teacher can still choose to use it. It’s just, rather than that being the default – “Everyone take out your laptops” – it’s now, “Okay. If  I wanna use tech, I gotta get my kids up, I gotta traipse ‘em across the school, that’s gonna take five minutes. I gotta sit down, I gotta get logged into their computers, that’s gonna take another five minutes. At least I have control over it.” Is it worth those ten minutes of learning time lost to use the tool? If the answer is yes, do it. But if the answer is no, and nine times out of ten it’ll be, “No, actually, there’s an analog way we could do it probably better,” then you won’t use it.

So it’s just about reintroducing friction to the decision to use tech again so that it becomes very specific. We are using it for purposes we understand, not simply because it exists right now. So it’s not about scrapping it, it’s about controlling it again.

[Kris Perry]: If tech is kept in the classroom to any degree, is there a useful model for integration that prioritizes learning, and are there certain ages or grades where you think that should start?

[Dr. Jared Cooney Horvath]: No. No, no, no, no, and no. I hear it. I wish it were gonna be that simple. I have yet to see a very valid, consistent use case for tech where it will always work with every kid everywhere if you do it like blank. There are some shining spots. If you use digital technology as a tutor to supplement learning, you seem to get good outcomes. But there are very big caveats to that in terms of you have to test people on the system. If you test their skills off the system, they lose it. Also, the time spent on digital tutoring, if you did analog tutoring, they learn more.

So I mean, is it, like – I mean, I don’t know how you want to do it. So I would just say no. When it comes to what age, it’s less an age thing, more a skill and development thing. What are you ready to teach them? If you believe that they’re ready to learn how to type, then you probably are going to need a keyboard. That might be in second grade, that might be in fourth grade. If you decide that’s important for you guys, bring it on. If you think you need to teach them to code, whatever year that is, far be it for me to tell you when to do that.

The only thing I’ll say is remember your progressions. Always do the hardest stuff first and then move into the easiest stuff. If you’re learning to write, always learn to write by hand first. The nuance and complexity of handwriting is arguably the most complex motor skill human beings have. When you learn to do this first, and then you learn how to type, which is just a gross motor movement, very consistent, very predictable – this translates very well into this, and you learn to type quicker and you learn to type more efficiently. But flip it – if you learn the early thing first, typing, and now I try and backtrack and say, “Now can you learn how to do pen and pencil handwriting?” The two almost never correlate because now they have to build an entire new motor program. If you’re teaching kids math, teach them to do math first. When they’re good at math, then bring in the calculator. Always learn the hardest way first and then make it easier as you go along.

And so I think even with something like coding – I was talking to someone recently – I don’t know the first thing about coding, I always just assumed if you wanted to teach coding you needed a computer. And he said, “No, the best way to teach coding is analog. That’s how you learn to think through code. Then we bring in the computer once you have the processes there.” So even in that example, even with coding, the argument can be made that you don’t even need the tech there first. So whatever age bracket, when you want to teach things, that’s up to you. But always remember: go hard to easy.

[Kris Perry]: Reminds me of learning music or another language. There’s so many examples that would work well with that point. Excellent point.

Are you in touch with educators who are dealing with the influx of tech in classrooms? And what’s your read on the educator perspective on using AI in tech for learning?

[Dr. Jared Cooney Horvath]: I’ve been doing this for a long time. About 70% of teachers are completely done with all this stuff. They’re over it. They have been for a while. About 20% don’t care. It’s neither here nor there for them, and only about 10% love this stuff. But that 10% has the loudest voice. And this is where one of the big problems comes in is teachers, we just don’t have a voice, man. I mean, administrators have a voice, parents have a voice. Teachers don’t because our job is on the line. If we speak up and say, “I don’t want to do this anymore,” they’ll say, “Cool, get out.” 

Like, I worked with a district in Kansas where every kid, they had a one-to-one laptop program, every laptop was feeding data to the district, and teachers were mandated to use the computer for a certain percentage of each class, otherwise they might lose their job. They’re timing me now for – okay, it doesn’t even matter if – that’s like people saying you have to use AI 90% of your job. Why? Why don’t you let us decide if it’s meaningful or not instead of, “We’re gonna tokenmaxx this stuff?” 

I’d say teachers, by and large, educators are done with this. But we just don’t have a place to push back. And that’s where parents come in strong, is when parents start to push back, administrators have to listen. And when parents push back, your best allies are gonna be teachers; they will never lead the charge, but my goodness, they are gonna be that second wave of the charge that’s gonna get everything done. But it’s up to parents to make that first kind of charge.

[Kris Perry]: But then the parent is in this position of having to work through the policy process. Often they can’t go directly to the teacher and say, “Hey teacher, can you stop with the laptops and the apps?” It becomes this, you know, larger problem, right? A systems problem, a triangulated problem where the teacher and the parent don’t really get to decide. It’s being decided at a level beyond the classroom that requires the parent then to find the time and the energy and the will to show up at a school board meeting or at the principal’s office or, you know, all the places you go to say, “I don’t really like what’s happening. It doesn’t seem like it’s good for the teacher or the student.” But I love the idea that parents are starting to see that this isn’t working well at home or at school and they’re probably inadvertently, indirectly helping teachers.

One thing we wonder about at the Institute is whether or not tech use is affecting rates of teacher burnout. You say they’re over it, 70% are over it. Is that because they’re burning out or because they recognize it’s limited utility?

[Dr. Jared Cooney Horvath]: Yeah, I’d say a little bit of both. It’s become the bugbear in a lot of cases, where it’s – it has to be there, it has to be used, and it just isn’t great. And so it’s the same with all tech. Remember, like there’s not a single piece of tech that’s come in that hasn’t claimed, “We’re gonna make your lives easier. You’re gonna have to do less work.” But when you actually track the data and research, no, it leads to more work. Emails didn’t lessen the time you talked to people, it increased the time you had to talk to people. AI didn’t lessen the time you coded, it made it so you had to pump out more code.

It’s the same thing here. Tech was, quote unquote, “supposed to” make teaching easier so you can focus on the good stuff. “We’ll handle the crap.” But you don’t know how much upkeep the crap takes. And so now I’m just blowing so much time. And so now most teachers are like, “Man, with digital LMS systems, kids have access to me 24/7. I’m fielding questions for my principal on weekends. I’m grading all night long. It’s made work worse.” So I think that is driving that, as well.

And – but, I mean, another big aspect of that is going to be the children, is the students. In a lot of schools the behavioral problems are through the roof because executive functioning is way down. One thing tech doesn’t teach you how to do is how to handle this, how to handle your emotions, how to handle your thinking. Anyone who’s a parent, who’s ever had to take the TV away from the kid, you have about a 30 to 60 minute window where they’re gonna be nuts till you get that out of their system to bring them back online. It’s the same thing in school. And I think a lot of teachers are just fed up with the lack of respect and lack of behavioral maturity in schools these days.

[Kris Perry]: We could talk for hours. We didn’t talk at all about dysregulation and all of these sort of more social-emotional impacts of AI and computers and having them on in the classroom and at home and, I mean, it is multi-layered to be sure. Is some of that, all of the complexity of this, how you got yourself from being a teacher to being a cognitive neuroscientist? What was the “aha” for you?

[Dr. Jared Cooney Horvath]: I just – see, I was doing this before tech was the big – the focus when I was teaching was the brain, was neuroscience. It was the decade of the brain. So brain books, brain gym, brain labs, everything was brain. And so that’s where I figured, “All right, cool, I’m gonna go back and learn all this stuff.” And tech was always tangential in the sense that if you took a master’s course with me about learning, we only talked about tech for maybe an hour. It’s at the very end, when you know about learning, I could say, “Okay, I want you to consider digital technologies: good, bad for learning? Why good, why bad? Link it together. Go.” And that’s about all it takes, is about an hour to discuss it. And most people are like, “Yeah, well, now that we understand learning better, it’s probably a pretty weak tool for that.”

So I had no intention of ever doing the tech talk or the tech games. It just so happened it’s just – it’s always bleeding there. And then COVID hits and everyone goes tech and then COVID stops and no one pulls that band-aid off. They double down on it. And that’s when it’s like, “All right, let’s lock it down.” Someone’s gotta tell everyone what’s happening and hopefully link that back to learning in a positive way that says, “I’m not just mad at tech. There is a way to learn, and we’re just not doing it right now.”

[Kris Perry]: I couldn’t agree more. Your congressional testimony about ed tech in schools went viral earlier this year and it seemed to really strike a chord. We watch experts all the time here share their insights, and it’s often hard for science-based perspectives to break through into popular culture and help bring visibility to an issue. What is your secret? How do you break through? How do you bring all of this evidence into a way of thinking and sharing that is easier for all of us to digest?

[Dr. Jared Cooney Horvath]: I wish there was a secret. It’s just work, work, work. The – I will say, when it comes to translating science into practicalities, I’m very lucky because I started as a teacher. I know what teachers want to hear. I always say, whenever I’m thinking about neuroscience, I say, “What would I have wanted to read as a first year? What would have helped me when I was a first year in a classroom?” And that’s how I kind of frame everything. 

But beyond that, it’s just – you keep chugging. I mean, I’ve been doing this for well past fifteen years now, and that clip is the first one that anyone actually pays attention to. I’ve got hundreds of clips online that ten people have watched that I think are millions of times better than that. It’s just that one hit the right chord at the right time. So believe it or not, I don’t put much stock in that. I just put more stock in, “Do good work. Just keep doing good work. Don’t try and chase popularity. Don’t try and chase the big movement. You do good work, one teacher at a time, one school at a time. That’s how we’re gonna make our move.” Every once in a while something like that pops. Smile, say that was fun, and then get back to doing good work.

[Kris Perry]: What do you think parents or educators can do to counteract the seemingly overwhelming amount of tech in today’s classroom?

[Dr. Jared Cooney Horvath]: From a parent angle, in your household, learning becomes analog. Buy a printer. Anything that has to do with learning in your house, print it out. If it’s a reading your kid has to do, print it out. Worksheets, print it out. Essay, have them write it by hand. As much as you can, you do analog in your household. Now, beyond the learning stuff, now it’s up to you and your family. How you guys want to do smartphones. If you want to do a Sunday tech free day, that’s all up to you. But when it comes to learning, as much as you can, you get them offline for that learning. And there will be times where teachers are like, “You have to submit your paper via this portal or you can’t.” Well, watch when your kid hands a paper to that teacher. The teacher’s not gonna crumple it up and throw it away. They’ll find a way to get it in there. But, if for some reason they don’t, then you submit it. You say, “Give me your handwritten work, I’ll take a picture, I’ll submit it. Cool.” 

In the classroom, then, from a teacher aspect: again, man, we – our jobs are on the line, so we’re very, very limited in what we can do. But you know if you’re a teacher, your room is your room. It’s called “the black hole of the classroom.” When that door closes, you can do what you can do. And that means in your classroom, you can pull tech back significantly. And you can start to track data in your own classroom. And if you can come out at the end of a term or a semester and say, “By the way, we used textbooks, maybe no one even knew. And you see how my kids did a little bit better than everyone else,” or, “You see how they did better than last year,” or, “The data shows that my kids were a little more engaged than everyone else.” If you can just do you in your classroom and just collect some sort of data so you can show other people that what you’re doing is working, that’s how we start to get more people on board. 

But I’m shocked, yeah, how many schools where it’s the teacher doing it quietly. And I know that’s a term AI uses, so we’re not supposed to, but they’ll just do it quietly. They just do it themselves, collect their own data so that at the end of the year everyone goes, “Well, wow, you were so good at teaching. What did you do differently?” “My kids took notes with a pen and paper. That was it. That was the only thing I did that no one else did this year. And it seemed to work.” That kind of thing. So as much as you can, you do what you know is right and then try and get some data so that you can prove it to other people when they come calling.

[Kris Perry]: Jared, thank you for joining us and sharing your insights. This has been an incredible and fascinating discussion, and we’ve discussed today conversations about technology and education are often framed as a choice between embracing innovation or resisting it. But the more important question is how we ensure that technology serves children’s learning and development, not the other way around. 

For listeners interested in learning more, we’ll include links to the show notes to Dr. Horvath’s work, as well as Children in Screen’s resources on AI and education, school technology, and healthy digital media use. If you enjoyed this episode, please subscribe to Screen Deep wherever you get your podcasts, leave us a review, and share this conversation with a friend, colleague, parent, or educator. To stay up to date on the latest research and practical guidance about children and digital media, visit childrenandscreens.org and sign up for our newsletter. Thanks for listening, and we’ll see you next time on Screen Deep.