Rewind: How AI is fueling an existential crisis in education

Hey, everyone, Nilay here. We’re off today, while the team and I are cooking on a lot of really great stuff in the coming weeks. We’ll be back with an all-new interview on Monday. In the meantime, we really wanted to highlight this episode we first aired in the fall, because it’s about a huge subject: AI in schools. The school year is starting to wrap up now around the country, and we’re no closer to figuring out how to thread the needle about generative AI in education than we were in September. Links: A majority of high school students use gen AI for schoolwork | College Board About a quarter of teens have used ChatGPT for schoolwork | Pew Research Your brain on ChatGPT | MIT Media Lab My students think it’s fine to cheat with AI. Maybe they’re on to something. | Vox How children understand and learn from conversational AI | McGill University ‘File not Found’ | The Verge Subscribe to The Verge to access the ad-free version of Decoder! Credits: Decoder is a production of The Verge and part of the Vox Media Podcast Network. Our producers are Kate Cox and Nick Statt. Our editor is Ursa Wright. The Decoder music is by Breakmaster Cylinder. Learn more about your ad choices. Visit podcastchoices.com/adchoices

• 0:00Speaker 00

  Company spend can be complicated. Managing it shouldn't be. AMEX Corporate connects cards, expenses, and payments all in one place. That's less time managing work and more time doing work that matters. Built to simplify, designed to scale. This is AMEX Corporate. Get started at go.amex slash AMEX Corporate. Terms apply.

• 0:29Speaker 01

  This episode is brought to you by Michelob Ultra. Good times mean competitions between you and your friends, like seeing who can throw some pigskin the farthest or rush the field the fastest. Winner gets an Ultra. Because it's not just a superior-like beer, it's also giving fans superior access to their favorite sporting events. Superior is worth playing for. Visit michelobultra.com/superioraccess to enter. Enjoy responsibly. Copyright 2026 Anheuser-Busch InBev Michelob Ultra Beer, St. Louis, Missouri.

• 0:58Speaker 02

  Are your ad campaigns lighting up the dashboard but not the pipeline? That's bull spend, and marketers are calling it out in. Dashboard, confessions.

• 1:07Speaker 03

  My boss asked for results, so I opened my dashboard for the only positive-sounding metric I had: impressions.

• 1:14Speaker 02

  Cut the bull spend. See revenue, not just reach. LinkedIn delivers the highest return on ad spend of major ad networks. Advertise on LinkedIn. Spend $250 on your first campaign and get a $250 credit. Go to LinkedIn.com/campaign. Terms and conditions apply.

• 1:27Speaker 04

  Hey, everybody. It's Eli. The cutters are off today while the team and I are cooking up a lot of really great stuff for the upcoming weeks. We'll be back with an all-new interview on Monday. In the meantime, we really wanted to highlight this episode we first aired back in the fall because it's about a huge subject: AI in schools. The school year is starting to wrap up now around the country, and we're no closer to figuring out how to thread the needle about generative AI in education than we were back in September. Lots of people are worried about students using ChatGPT to cheat on assignments, and that is a problem. But really, the issues go a lot deeper, to the very philosophy of education itself. Dr. Adam Dubay, an expert in educational technology from McGill University, joined me on the show to talk through how generative AI fits into education right now and where it might be heading in the future. We also talked to a whole lot of actual teachers.

• 2:12Speaker 04

  You'll hear their voices throughout this episode, and we kept hearing one thing over and over again: What are we even doing here with AI? What's the point of this? It's a big question with not a lot of answers. Here it is: AI in education. Enjoy.

• 2:27Speaker 04

  Hello and welcome to Decoder. I'm Neelai Patel, editor in chief of The Verge, and Decoder is my show about big ideas and other problems. We've talked a lot about generative AI on the show lately, which is a very big idea that is causing quite a few problems. And one thing we keep hearing about over and over again is that generative AI is causing a lot of problems in schools. There are a lot of people out there, including many of the listeners of this show who email us, who are worried about the obvious problem: students using ChatGPT to cheat on assignments. But when our team went and poked at the story, they found that the issues in education with AI go a lot deeper, to the very philosophy of education itself. We sat down and talked to a lot of teachers. You'll hear a lot of their voices throughout this episode. And we kept hearing a common theme: what are

• 3:11Speaker 04

  we even doing here? What's the point?

• 3:14Speaker 05

  Hi, I'm Evie May. I'm an instructional designer at a small college in Michigan. When I attended the Online Learning Consortium's Innovate Conference in 2024, one of the presenters discussed using various gen AI tools to give feedback on student papers. So if this technology becomes more ubiquitous, we'll have courses created by AI, graded by AI, with submissions from students absolutely generated by AI. So it begs the question, what are we even doing here in higher ed?

• 3:48Speaker 04

  Now, every teacher is having a different experience with AI in the classroom and with their students. But the common thread is that so many of those experiences feel bad. A few teachers who talk to us find tools like ChatGPT are helping their workflow. But a lot of others are facing those deep existential questions like you just heard from Evie. Luckily, there are experts in education and educational technology who research what's going on in a more detailed way. So I sat down with Dr. Adam Dubey from McGill University to talk about how generative AI is fitting into education right now and where all of this might be going in the future.

• 4:36Speaker 04

  Dr. Adam Dubay, you are Associate Professor of Learning Sciences at McGill and co-lead of the McGill Collaborative for AI in Society. Welcome to Decoder.

• 4:44Speaker 06

  Thanks for having me here, Neil.

• 4:46Speaker 04

  There's a slide you have about the lessons we have learned and not learned from the Internet and mobile, and it says, "Digital natives do not exist."

• 4:53Speaker 06

  Yes.

• 4:54Speaker 04

  Can you briefly explain what you mean by that?

• 4:57Speaker 06

  The term "digital natives" was coined back in the early 2000s, and it was this idea that perhaps kids that are born into technology understand it better than, say, what he called "digital immigrants." And there's been a lot of problems with this type of language and the framing about how he talked about it. And then there was 20 years of research to see if this is actually true. Are kids that are born, growing up with technology, better at using it than people that adopt it later on in life? And the research for 20 years has shown this isn't actually the case. It's not that because you're young and you grew up around it. It's just about how much you've used it, how much exposure you've had to it. And this really matters when we talk about AI in education and technology in education, because even though a kid grows up using YouTube

• 5:40Speaker 06

  or a phone for playing Roblox, doesn't mean they know how to use technology to actually learn, but we assume that they do. Teachers assume that kids know how to use technology in the classroom. They assume they know how to use it for learning purposes because they use it for YouTube. And so, we aren't digital natives. We just have previous experience using technology for specific things. And it's caused a lot of problems when it comes to education because we assume kids have skills that they don't.

• 6:08Speaker 04

  We've done a lot of coverage over the past five years here at The Verge about what smartphones and tablets have done in education, in particular, as it relates to really core computing concepts. One of my favorite stories we've ever done is called "File Not Found", and it's just about kids in STEM who don't know how file systems on Windows work. And so, the STEM professors in college have to spend a day explaining what files and folders in Windows are before they can go use the radio telescope or whatever tool they need to use next. And that always felt to me like... We take for granted that the frameworks of the past will be intuitively understood by the kids in the future, but those frameworks change. Would you put AI into that kind of

• 6:52Speaker 04

  category that this is a framework change for how we use computers and that frameworks of the past might just be abandoned?

• 6:58Speaker 06

  I think the way that we interact with computers could be changing by having us engage with it through natural language interfaces. Unfortunately, the logic that underlies that system is still really important for being able to interpret the answers that it's giving us. And so, yes, kids, people growing up using a computer where it's say a primarily a text-based or a voice-based system, they're not going to think of it the same way than someone who grew up with a file system and engaging with individual applications instead of everything being launched through ChatGPT, say for example. That's going to be a problem not for them using and interacting with the system and asking it to do things, but then how they actually interpret the way those systems

• 7:40Speaker 06

  give them answers and how they evaluate it. Can they actually make sense of these, the responses and then make critical judgments about it? And so, actually, this is an area of research that myself and my PhD candidate Nundan are working on where we're looking at children's theory of artificial minds. We're trying to understand how do children think computers think? Specifically, how do they think that AI reasons, if we can say that it reasons? And then what's the impact that that's going to have on how they learn from AI that's put in their schools and in their homes, like smart speakers that are already everywhere? And we're just starting these types of studies, but these devices are now being deployed actively into schools where we don't have a great understanding of this yet.

• 8:22Speaker 04

  That idea that we don't really understand AI yet, that a lot of people don't know how it works and that we have no long-term data about its effects in the classroom because it's so new. Well, that's a really big point of contention that we heard from a lot of teachers.

• 8:34Speaker 07

  I'm recently retired high school English teacher Anne Lutz Fernandez. During my last year of teaching, I began to see more students using generative AI to replace their own reading, thinking, and writing, even creative and personal writing. We're treating children like guinea pigs on an untested and unproven and unregulated host of products. It feels to me like we haven't learned some key lessons, a lot of them very recent. One of those during the pandemic was the costs of unhuman teaching and learning.

• 9:09Speaker 07

  And I worry that as we did with cell phones and over reliance on one-to-one devices, we're going to wake up a decade or more from now and realize we jumped on a tech bandwagon that keeps kids tethered to screens, harms them, and harms learning.

• 9:26Speaker 04

  That shift to personal one-to-one devices was really huge. And it means that there are a lot of screens in K through 12 education now. And it feels like there are some lessons we should learn about how those prior technologies were introduced.

• 9:39Speaker 06

  I was a math cognition researcher that looked at how children understand simple things like learning how to add. And then I got into studying technology because Apple came out and was pitching the iPad as the future of education. And then a bunch of math apps were launching that were saying, okay, this is how your child's going to learn math the best. We were actually testing this. Early on, back in 2011, 2012, we were giving kids a bunch of different iPads with learning apps. And you would think that the kids knew how to use them, but 90% of the interactions they had with that learning app were actually complete mistakes and errors. They were just randomly tapping around the screen. And so there was a lot of guessing, but the apps actually had no negative consequences for getting stuff wrong. So we called this as just being the app was too dumb to cause

• 10:24Speaker 06

  a mistake. It didn't matter if you interacted with it in a random way. It always just progressed. So this sort of idea that kids get how to work with technology is actually a byproduct of an oversimplified design of a lot of the apps that are used for learning. And then now we've got new systems coming in where it seems like, well, kids can just talk to AI speakers and they can just talk back to them. And then example of this is that video that Sal Khan put out with Migo Tutor. It's like, well, look, I can just sit my kid down in front of this. I can say, teach them how to solve this math problem, and it just does the thing. Okay, but is that child actually benefiting from that experience? Are they interpreting the lesson correctly? Is that going to help them understand it? Is that

• 11:05Speaker 06

  a good way to teach whatsoever? But because it looks so easy, we get convinced that this is somehow useful. And I think that's going to be a problem right now with the generative AI tools that are coming out and the way they're being pitched.

• 11:17Speaker 04

  So kids aren't predisposed to using iPads or AI any better or more competently than adults are. In fact, they might be worse at it if they don't have the experience they need to be able to tell how AI is really working. But even still, research from Pew in January found that about a quarter of teens were already using ChatGPT in their schoolwork.

• 11:37Speaker 04

  That number edges up to a third for high school juniors and seniors. In May, the College Board published research saying 84% of high school students were using some kind of generative AI tool in some way for schoolwork. Does that match up with the numbers you're seeing in your research?

• 11:52Speaker 06

  I looked to research by Victor Lee. He looked at 4,000 high school students. And what matters, not just if they're using it, but what they're using it for. And so the biggest use of them is that students in high schools are using it to explain concepts to themselves. So that's 80% of them. And that's gone up from a year previously. They're using it to generate ideas for assignments. And that's about 70% of the use cases. And then they're using it to summarize texts instead of reading it. 40% are using it to edit portions of text. And then at the very bottom, and I think this is something that probably people are going to disagree with, is that only 10% of students are actually reporting that they're using it to generate their whole assignment, which is what people are really worried about when they think about AI cheating in schools.

• 12:37Speaker 06

  But that number is pretty consistent and hasn't really changed over the years. The percentage of students that report actually cheating has stayed around 10%. It's just how they cheat changes over time. So paying someone or copying stuff off the internet and then now. And so the numbers that you say reflect what we see, but then we're like, how are they using it? They're using it in different ways for different purposes. And then we can debate whether or not these different uses are even good whatsoever. Obviously cheating is bad, but is it good at summarizing text? It's not, but you know, that's where we can get more nuanced questions about like these different uses. Or is that actually going to be beneficial for students learning?

• 13:14Speaker 04

  Can I just ask about the cheating number specifically for one second? Are you just finding that no matter the technology, 10% of students are dumb enough to self-report themselves as cheaters? Is that, is that what that demonstrates?

• 13:26Speaker 06

  I like the, I like the framing. There is dedicated researchers and institutes that look at academic misconduct and they actually do a lot to get students to be honest in their reporting within this research. And that's where that number comes from. I have seen students openly admit to using generative AI in open symposium in front of their professors. So some people are perhaps not that bright when they did this. But here, that 10%, I think what really matters is that we're trying to see, okay, what is the real prevalence of this in the student population? And people have dedicated themselves to studying that and trying to find a way to have students be honest about it. And so in that way, we really know what the problem is.

• 14:07Speaker 04

  I mean, if you told me that 10% of teenagers were self-destructively stupid, I would just believe it, no matter what the data showed. I was in that 10% for sure.

• 14:16Speaker 06

  On that note, we have to take a quick break. We'll be right back.

• 14:28Speaker 01

  This episode is brought to you by Michelob Ultra. Good times mean competitions between you and your friends, like seeing who can throw some pigskin the farthest or rush the field the fastest. Winner gets an ultra because it's not just a superior light beer. It's also giving fans superior access to their favorite sporting events. Superior is worth playing for. Visit michelobultra.com/superioraccess to enter. Enjoy responsibly. Copyright 2026 Anheuser-Busch InBev Michelob Ultra Beer, St. Louis, Missouri.

• 14:57Speaker 02

  Are your ad campaigns lighting up the dashboard? But not the pipeline? That's bullspend and marketers are calling it out in. Dashboard Confessions.

• 15:06Speaker 03

  My boss asked for results, so I opened my dashboard for the only positive sounding metric I had. Impressions.

• 15:14Speaker 02

  Cut the bullspend. See revenue, not just reach. LinkedIn delivers the highest return on ad spend of major ad networks. Advertise on LinkedIn. Spend $250 on your first campaign and get a $250 credit. Go to LinkedIn.com/campaign. Terms and conditions apply.

• 15:27Speaker 08

  College football is back. So Hilton called to me, the superstition concierge, to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky jersey? Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition. Even a marching band wake up call at 5:55 in 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton, for the stay.

• 16:01Speaker 04

  Welcome back. I'm talking with Dr. Adam Dubay about what his research is saying about generative AI in schools. Before the break, we were talking about how all of this is just new technology. And as a result, it's kind of a mess. Students are using it to cheat, although maybe not as many as we're worried about. Teachers are feeling pretty confused about how to respond. And there's just not a lot of clarity from anyone in response.

• 16:24Speaker 04

  That kind of usage is leading to pretty whiplash policies across schools at every level. There's the we're going to ban it entirely kind of movement. The schools in my kids' district, they've just fully banned smartphones from schools. That's here in New York. That's statewide. There's we have to put AI everywhere to get these kids ready. There's Salk on saying, just let my robots teach your kids. This is a pretty wild mishmash of policies and approaches. What is the general shape of it that you've seen?

• 16:52Speaker 06

  It is very fractured. And it depends on who the leader of that school system is and on their view of technology, and then on the broader community around that school. The parents in that community, do they have a negative attitude towards technology? Right now, there's a big anti-screen movement that's happening. We see cellphone bans, concerns about social media from parents. This is increasing. You have the larger community influencing the way that school leaders think about technology. But then you've got some school leaders who are saying like, okay, we're resource constrained, our budgets are being cut, and they're seeing technology as potentially a way to save money. And so they're turning to generative AI as a way to maybe make up for not having

• 17:34Speaker 06

  enough educators in their classrooms. Or maybe they truly believe that it's a transformational tool. But you can't, there is no one consistent system. It varies almost from school district to school district. And I've spoken with school leaders across our provinces because we run education at a provincial level. Now, there's no federal sort of oversight. All the principals complain that there's no overarching guidance, that everyone has having to figure it out by themselves. And that leaves it up to the factors at the local level influencing whether or not AI is seen as a potential positive or negative.

• 18:08Speaker 06

  Or not, it's a positive or negative for teachers, the admin, or for students. There's also differences there. A lot of teachers think students shouldn't be using it, but it's okay for them to use it. Or the admin thinks this is going to help us save time for teachers marking students' assignments. So we can save some money there, but we don't want our students using it. But we're going to use it to analyze student data. Right? So there's even a mismatch and a disagreement within schools about the role of generative AI. Right now, these systems are being sold to educators to generate lesson plans, to evaluate student work, to do learning analytics. And if you're having this deployed in your school and before you teach a class, you're being told, it was like, okay, well, we're going to cut back on how much teacher preparation time there is.

• 18:53Speaker 06

  But we bought Magic School for you. And so it's going to generate a lesson plan for you. So it's, don't worry, you're going to have plenty of time to do it. It's like, actually, keep track of how much easier it is to generate your lesson plans and do your work with these tools.

• 19:08Speaker 04

  A few of the teachers we spoke with really were excited by the idea that generative AI could be a time-saving tool and actually help them out when it comes to managing a busy workload with too few resources.

• 19:18Speaker 09

  I'm Paul and I teach middle school science in Raleigh, North Carolina. And the thing that has me most excited about generative AI technology is the way that it unlocks teachers' ability to do better teaching in ways that many of us really want to. We're constantly being told about new research that shows that there are better ways to teach. But many of these strategies and techniques, they require a lot of time and effort for us to like, learn more about them and to build content with them. By partnering with an AI tool like ChatGPT, a lot of this becomes way more doable. And so I find that I'm able to integrate better strategies into my teaching because I know that I have support when it comes to building new materials with those strategies highlighted.

• 20:03Speaker 04

  That's all pretty interesting, but Paul's position is part of a distinct minority. Certainly, at least amongst the teachers who spoke with us. Here's Evie May again.

• 20:11Speaker 05

  Despite many attempts to incorporate it into my workflow, I've found that gen AI is more trouble than it's worth. And that's beyond the simple fact of the technology's unethical plagiaristic roots and environmental destruction. Just purely on a utilitarian level, I can do better work much faster when it comes to designing course materials. At most, I would use ChatGPT to clean up auto-generated YouTube captions, but YouTube's already improved this on their own end, so it's kind of a moot point.

• 20:44Speaker 04

  And then sometimes, as some teachers told us, generative AI can make things actively worse than they were before.

• 20:50Speaker 10

  My name is Anne Rubenstein. I'm a historian and a professor of history at York University. One of the things that I do as a scholar is I help prepare collections of documents from the past on specific topics that are then published as part of a digital history project that goes out to university libraries mostly. Because I am a historian of Mexico, the documents that I'm preparing for them are Mexican, and they're in Spanish. The publisher decided that we should, along with providing the original documents, we should provide translations into English, since that's the language that the majority

• 21:33Speaker 10

  of people using these teaching tools are going to be comfortable with. So, great. I said, "Great, I've got a friend who's a translator. We'll get them to translate these documents, no problem." And they said, "Oh, no, no. We've bought new software that will translate for us, and we don't need to go to the expense and trouble of hiring a human translator, because this translation software is going to be great." And I was skeptical, but I said, "Sure, let's try it." And so, we tried the software, and here's what it did. It hallucinated. It made crap up. It inserted entire sentences, and in a couple of cases,

• 22:15Speaker 10

  entire paragraphs, into the document that did not exist in the original. I, if you don't understand why that is a very, very big problem in a collection of translated primary source documents for history students, I invite you to come take some history classes, and then you'll understand why that's an enormous problem. Luckily, the publisher also understood this was an enormous problem. So what they decided to do was hire a translator, whose job it was to go through these machine translated documents and restore accuracy and clarity to them.

• 22:52Speaker 10

  And that ended up costing just about twice as much as just hiring a human translator would have done.

• 22:59Speaker 04

  In a strange way, it might help when the hallucinations are incredibly obvious, because then you can tell that the tool isn't working for you. But sometimes it can be a lot harder to spot if a tool is actually saving you time or improving your work when you first start using it, and then generative AI produces polished content and answers to questions so quickly that it feels like it's giving you something meaningful.

• 23:18Speaker 06

  Is it actually saving you time? I speak to teachers and they say like, well, I use generative AI and it helps me generate my lessons. It helps me write emails. Actually monitor and try to keep track of, is this actually speeding things up? There's a lot of research that shows that, say, for example, with coders, they actually end up being slower when they use these systems because they have to fix all the mistakes. And I think that we might see a similar pattern that happens with educators. They're using these tools to be more efficient, they think. But if they actually tracked how long it takes them to generate a lesson versus how long it takes to fix the lessons that generative AI produces, it might not actually be faster. I see some educators that are enthusiastic about the time-saving it can give them, but I'm not sure it's actually saving anybody's time.

• 24:03Speaker 04

  So from the teacher's perspective, generative AI in schools is a workplace issue, a labor issue. And there's a lot of research out there, both some older research and also some new research recently published in the Harvard Business Review, about how workers feel when they're forced to use specific tools or behave in certain ways that devalue their own expertise or autonomy. How's that working out for teachers?

• 24:25Speaker 06

  There's some research that looks at school climates and teachers who get demotivated for their use of generative AI in education, and what causes demotivation. And for them, it was being forced to use these systems when there was a top-down rule that you had to use generative AI, maybe for lesson planning or writing emails, or if we're doing student feedback. That is demotivating for educators. They don't like being told which tools to use because it feels like it's removing their autonomy. And so whenever we remove workers' autonomy or their own scent, basically their control over their own work environment, people get demotivated. So it's not surprising that workers feel demotivated when generative AI is being forced into their workplace, because they have less of a say on what they

• 25:09Speaker 06

  get to do. And as human beings, we want to be creative. We want to produce. We want to feel like we have a control over our lives and our work. And then something comes in and takes control away. We're not going to like it.

• 25:21Speaker 04

  We have to pause here for a short break. We'll be back in just a minute.

• 25:33Speaker 01

  This episode is brought to you by Michelob Ultra. Good times mean competitions between you and your friends, like seeing who can throw some pigskin the farthest or rush the field the fastest. Winner gets an ultra because it's not just a superior light beer, it's also giving fans superior access to their favorite sporting events. Superior is worth playing for. Visit michelobultra.com/superioraccess to enter. Enjoy responsibly. Copyright 2026 Anheuser Busch InBev Michelob Ultra Beer, St. Louis, Missouri.

• 26:01Speaker 02

  Flowing ad budget on metrics that look great till the CFO sees them. That's bullspend, and marketers are calling it out in Dashboard Confessions.

• 26:11Speaker 03

  I remember telling my boss, "It'll be good for the brand," when leads were slow. Yeah, it wasn't.

• 26:18Speaker 02

  Cut the bullspend. LinkedIn lets you target by company, job title, and more. Advertise on LinkedIn. Spend $250 on your first campaign and get a $250 credit. Go to LinkedIn.com/campaign. Terms and conditions apply.

• 26:32Speaker 08

  College football is back. So Hilton called to me, the superstition concierge, to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky jersey? Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition, even a marching band wake-up call at 5:55 in 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton, for the stay.

• 27:05Speaker 04

  Welcome back. Before the break, we were discussing how generative AI is affecting teachers. For school administrators and people in the classroom, generative AI is a workplace issue. But the really big question, the one everyone is concerned about, is what these tools are doing for students. As we've discussed, there are studies that say generative AI in the workplace can actually demotivate adult professionals in a lot of ways. But anyone who's watched a kid really engage with ChatGPT's voice mode or Google Gemini can see that a conversation, no matter how one-sided, can really keep a kid's curiosity going when they fall down that rabbit hole and maybe even teach them something. So could there be some upsides to generative AI chatbots when it comes to learning?

• 27:46Speaker 06

  It does seem that using these tools can increase affect and motivation when it was designed to do that. Why is that happening? Is it the way these things talk to us? They congratulate us on our questions. They always provide an answer. It's a very positive experience. So maybe that sort of paradigm is causing the increased use of these systems, which is creating a loop of query and response and query or response. And that makes a lot of sense from a design standpoint. If you're trying to make a system where you want people to use it more because you're trying to make the case that you need the data centers and everything else in the use levels. It's like, well, it's positively reinforce questions and answers and make it a very rewarding experience to ask us things.

• 28:29Speaker 06

  But at the same time, it's always positively providing you these answers, but sometimes it's wrong. Well, what if they changed it a little bit and they said, okay, well, you know, I can't actually answer that question for you because I'm not sure if it's right. What if it was truthful? Well, then people might stop using it as frequently. And then maybe it's not as engaging and motivating. And so the question is, are do we think as an educational tool, it's good to somehow generate curiosity if the thing is lying to you about the information? It's like, would that be okay if it was a human being? It's like, well, here's our teacher, Jerry. He's in the class. He always lies to the students when they ask him questions, but he constantly, but he gets them going.

• 29:10Speaker 06

  It's like, well, that seems like a really weird position to take when we translate this over.

• 29:16Speaker 04

  There are some core skills that you should definitely have to learn. I actually think math is one of them, but you know, I do have friends who are like, screw it, I have a calculator. I can literally Google the answer to unit conversion and I will never think about it again. ChatGPT is that on a massive scale, right? You can just hand over some amount of skills to this robot about thinking about a lot of things. And maybe you'll just not be motivated to learn those skills because you know, there's a backstop. Whether or not the backstop hallucinates, you'll know there's a backstop and you'll never be motivated to learn those skills. Have you seen that dynamic play out?

• 29:48Speaker 06

  The classic example, as you said, is the calculator. And some people say that it didn't matter that we put calculators in classrooms. I actually had a colleague, Joanne Lefebvre, that looked at this throughout the 90s and 2000s in Canada, and we actually saw a decrease in math scores directly correlated with increasing calculator use, because when you're using a tool to do thinking for you, you're not practicing and actually encoding the information well enough to actually recall it later on. And so it's not surprising that when people use generative AI to do work for them, that they're not able to do it independently. Being able to store information in your memory requires effortful practice. It requires effortful memorization. It requires reflection. It requires thinking about, okay, what am I trying to understand?

• 30:32Speaker 06

  Connecting that to my other understanding. That's what builds a strong knowledge network. And when you use systems that just generate answers on your behalf, you don't engage in those practices. It just gives you the response and you passively consume it, and maybe you don't reflect on it. So it's not surprising that with the use of generative AI, one of the big effects that we see is that people are able to produce maybe work that looks more polished, but they don't remember the work that they actually wrote. That MIT study is the example that a lot of people have heard about, where they had people writing essays with or without ChatGPT or Google, and students had very poor memories for the essays that they wrote using ChatGPT. Well, that's because they actually weren't reflecting on their writing.

• 31:15Speaker 06

  They weren't engaged in the work that it takes to actually form substantial memories that you can remember later on. Now, should we care? Like, that's the thing, like, who cares? If you can produce the product, and if you can produce the work in the end, it's like, it doesn't matter. Well, if we think of the future down the line, where if you're using these tools to produce a piece of work, who is actually able to evaluate whether or not that work is any good? If everyone is just in the same level of expertise of just, I've used all these tools to produce the work? You actually don't have the internal knowledge, you don't have the internal skill set to say, like, is this good writing? Is this a strong idea or not? You don't know anything about the literature. You don't know anything about...

• 31:57Speaker 06

  The area that you're actually studying because you just use these things as the reference. You know, the people always say, I don't need to count. It's like, oh, I can just use a calculator. Well, I bet that person doesn't have to do mathematics in their job every single day. Right? So they're not regularly having to do unit conversion on the fly. So it doesn't really matter because they don't use it. But in your profession, in your daily lives, access to information really does matter. If you have to grab something, you don't want to have to turn to an external tool to make a judgment in a moment. I shouldn't have to be talking to you in this conversation. And then having chat GPT open over here to ask it things so I can have a conversation with you. That wouldn't be an actual productive conversation. It would impede how well I interact with my daily lives and do my job.

• 32:42Speaker 06

  And so it does matter when we don't actually know things and have a knowledge base of which we work.

• 32:48Speaker 04

  So then teachers are left with the challenge of getting students not to let chat GPT do the thinking for them, at least not in class. Sometimes addressing it head on is the way to go. Here's Anne Rubenstein again.

• 32:58Speaker 10

  What I've worked out to do with the first year undergraduates especially is not so much to tell them that they can or can't use this stuff to help them in their process of becoming history students, but to think with them about the ways in which this particular kind of software is likely to lead to bad results for historians in particular. What I tell them is as historians, we have a social responsibility to get our facts exactly right. Not only to say exactly when the specific event happened, but if we're quoting to back up our assertions, which we frequently do,

• 33:40Speaker 10

  we have to say who was speaking and also how we know what words they said, and also to quote their words precisely in the precise order, not to add any, not to leave any out. And to say where and when this quotation was made and to put it in its context. Once we've sort of gone through that with the beginning history students, then I talk to them about ChatGPT and similar software and I say, okay, how does it work? And I pretend to be slightly more naive than I actually am. And I say, okay, explain to me how this works. And usually in any group of, say, ten or more undergraduates, one of them is going to have a very clear understanding of how this software works.

• 34:23Speaker 10

  So I get them to explain it to me and incidentally to explain it to themselves and each other. And we talk about how the software can't by its nature actually know a thing. What it can tell you is what order words are likely to be in. And if we're very lucky, what will happen in the classroom as we're figuring this out together, going over it together, is that someone will say, oh, so they're bullshitting. And I say, yes, that's bullshit. And then we talk about what bullshit is and why we want to avoid bullshitting and why bullshitting is sometimes useful and important in life, but historians aren't allowed to bullshit. We absolutely cannot bullshit.

• 35:05Speaker 10

  We are the only people in the world who are never, ever allowed to use bullshit. And so then the lesson is you can use ChatGPT and similar software for all kinds of things, but you cannot use it in conducting historical research or writing about history because it is the exact opposite of what historians are supposed to do.

• 35:29Speaker 04

  What Anne just described, being able to talk through the reasons why you would or wouldn't want to use a specific tool, is really important because ChatGPT is just that, a tool. And students who are under a lot of different pressures might reach for any tool they have at hand. And maybe, aside from that 10%, Adam told us about, we're happy to just cheat, maybe it's not hard to see what pressures students are reacting to. In a way, they're just behaving rationally inside of the system that they're a part of, and that itself is a kind of problem. Are the educational systems we've set up actually designed to prioritize learning as an incentive for the students?

• 36:02Speaker 11

  I'm Brian S. and I teach technical communications at a Midwestern Research One university. I have a lot of engineering majors in my classes and my job is to teach them how to speak about engineering things to non-engineers. The big effect large language models have had on my job and my students is that it's really forced me to recognize how differently I see what's valuable in the classes from at least some of my students. In my classes, I teach a lot of how to write things. Some of it is formatted, or more of it is about tailoring a message to an audience, translating concepts from expert to non-expert. But as with most writing classes, the grades are based on the finished product, the user manual, or the proposal, or the report. I use those to evaluate how well the students have internalized the tools I've been showing them how to use.

• 36:45Speaker 11

  L L Ms promise that they can create those documents without having to learn all those intermediate steps. And when they're being used by a person who already has those skills, they can take some of the grunt work out of it. My students don't have those skills yet, and if they lean on L L Ms now, they'll never develop those skills. But for a pretty good-sized portion of my student body, that's not a problem, because one, they have limited time and they have physics exams and so on, and two, because they get graded on the finished product. For me, the finished product doesn't really matter. I've read enough proposals for free student parking on campus for multiple lifetimes. But for them, it matters because it can affect their academic standing, potentially their

• 37:27Speaker 11

  financial aid, and they believe that it can affect their ability to get a job or an internship. The grade matters a lot more to them than to me, in other words. So it makes sense that if there's a tool that promises a product that will help them pass so that they can concentrate on the stuff that they feel is more important to their career, of course they'll think about using it. That's the real tension at play here. How do we convince students that the value in the class is learning how to do things when the thing we measure is the end product, especially when there's a tool that can take some of the pain out of producing that product?

• 38:00Speaker 04

  Learning isn't just about homework. Homework is useful practice. But our systems reward the end product over the process. They reward the completion of the homework over the actual learning. So how would we need to change that so that students don't want or need to outsource all of their thinking to AI? Here's one final thought.

• 38:18Speaker 12

  My name is Todd Harper and I am a professor of game design at the University of Baltimore. A thing I have observed in 15-ish years of teaching at the college level is how much refocus there has been on the product when it comes to course assignments, papers, presentations, whatever. Students are aiming for the grade because the grade is the thing that hooks into important metrics. It's the thing that hooks into whether they graduate or not. Sometimes it influences financial aid, et cetera, et cetera. And students are under tremendous

• 39:01Speaker 12

  pressure that affects how they approach their college education. My university largely has students who are focused on getting out of here and getting a job. A lot of them work full-time or are full-time caregivers or have some kind of equivalent everyday pressure on them. They're taking multiple courses at a time. They're trying to make it all work out. And if a tool comes along and says, "Oh, you got a paper due? Just plug the question into me and I'll give you a plausible sounding result." And then the student can be like, "Great. That's one thing I can check off the list so I don't lose my mind trying to

• 39:41Speaker 12

  be alive in 2025." I get why that would have some appeal, but pedagogically, educationally, we don't assign homework, papers, presentations, projects. We don't give those to students because we want something in return. The thing that they give to us is not the point. The point is that when they're looking up sources or drawing the art or creating the thing that they turn into us, they are exercising the skills and the learning that we want them to develop in our classroom, that they have come to our classroom to develop. And yes, we have to evaluate them and we can't be there for the process.

• 40:26Speaker 12

  So they do have to turn something into us, right? Like the product is how we evaluate the process through the result. But the process is the important bit. And if what's turned into me, you know, look, air quotes, looks right, you know, looks plausible, which LLMs could be good at. In fact, it's probably the only thing they're air quotes good at is making things that look plausible. But if the student didn't do it, if there was no process, then what are we doing here? No real learning has happened. All that's happened is that somebody ticked off a box on a to do list. And I think it hurts students when that happens. What we need is not more tools that produce product.

• 41:10Speaker 12

  What we need is fewer stressors, financial, cultural, social, whatever. What we need is less pressure on students so that they can actually do the things that they need to do to get an education.

• 41:28Speaker 04

  I'd like to thank the many teachers who spoke with us for this episode, especially the ones who are willing to be recorded. I'd also like to thank Dr. Adam Dubey for taking the time to join me and thank you for listening. I hope you enjoyed it. If you'd like to let us know what you thought about this episode, and I'm sure a lot of you do, you can email us at [email protected]. We really do read all the emails. You can also hit me up directly on threads or Blue Sky. Or you can leave a comment on our YouTube channel. You can watch full episodes at DecoderPod. We also have a TikTok and Instagram. They're at DecoderPod2. They're a lot of fun. If you like Decoder, please share it with your friends and subscribe wherever you get podcasts. Decoder is a production of The Verge and part of the Vox Media Podcast Network. The show is produced by Kate Cox and Nick Statt. It's edited by Ursa Wright. Our editorial director is Kevin McShane. The Decoder music is by Breakmaster Cylinder. We'll see you next time.

• 42:19Speaker 13

  New and exclusive holiday decor just launched online at The Home Depot. Check out the viral Grand Duchess Collection with smart holiday lights you can control from your phone. Easily customize the twinkle across trees, wreaths, and garlands to match your holiday vibe. Create stunning light effects, then switch it up with the tap of your fingers. Even the man in red will be impressed. Shop the holiday decor you want, online only, at homedepot.com.

• 42:49Speaker 14

  Where's your playlist taking you? Down the highway? To the mountains? Or just into daydream mode while you're stuck in traffic? With over 4,000 hotels worldwide, Best Western is there to help you make the most of your getaway. Wherever that is. Because the only thing better than a great playlist... is a great trip. Life's a trip. Make the most of it at Best Western. Book direct and save at bestwestern.com.

• 43:19Speaker 15

  Close your eyes, exhale, feel your body relax, and let go of whatever you're carrying today. Well, I'm letting go of the worry that I wouldn't get my new contacts in time for this class. I got them delivered free from 1-800-CONTACTS. Oh my gosh, they're so fast! And breathe. Oh, sorry. I almost couldn't breathe when I saw the discount they gave me on my first order. Oh, sorry. Namaste? Visit 1-800-CONTACTS.com today to save on your first order.

• 43:47Speaker 14

  1-800-CONTACTS.