Is ChatGPT Killing Higher Education?  ## Metadata - Author: [[Decoder with Nilay Patel]] - Full Title: Is ChatGPT Killing Higher Education? - Category: #podcasts - URL: https://share.snipd.com/episode/c960d733-75e5-463a-b651-707fae167712 ## Document Notes ## Highlights - **AI Is Reshaping Academic Work** - AI is enabling students to offload essays, homework, and exams and is changing what academic work looks like. - James Walsh argues this shift creates a 'cheating utopia' that challenges the purpose of higher education. Transcript: Speaker 2 What's the point of college if no one's actually doing the work? It's not a rhetorical question. More and more students are not doing the work. They're offloading their essays, their homework, even their exams to AI. Tools like ChatGPT, Claude, Copilot, they're not just study aids anymore. They're doing everything. We are living in a cheating utopia. And professors know this. They see it. It's becoming more common, but they can't always prove it. And more often than not, they're too burned out or unsupported to do anything about it. So what are we doing here? What is higher education at this point? I'm Sean Elling, and this is The Gray Area. My guest today is James Walsh. He's a features writer for New York Magazine's Intelligencer and the author of the most unsettling piece I've read about the impact of AI on higher education. Walsh spent months talking to students and professors who are living through this moment. And what he found isn't just a story about cheating. It's a story about ambivalence and disillusionment and despair. A story about what happens when technology moves much faster than our institutions can adapt. We cover a lot of ground here. We talk about the students who are cheating because they can, the ones who are cheating because they feel like they have to. We talk about the professors who are giving up, the administrators who don't want to confront the problem, and what all of this means not just for the future of college, but the future Of writing and thinking. ([Time 0:02:38](https://share.snipd.com/snip/153945a1-d040-45c5-a016-481ebf5e60c0)) - **Classroom Uses And Failures** - Students use ChatGPT for brainstorming, outlining, note-taking, summarizing textbooks, and coding assignments. - A Berkeley professor said many CS students use AI on take-home work and then fail in-class tests on the same problems. Transcript: Speaker 2 Well, let's walk through the tools that students are using, because I still think there are a lot of people who just aren't familiar with any of this, have not yet dipped their toes into The wonderful world of LLMs and chatbots. What are the main platforms? Are we talking about, you know, ChatGPT, Clog, Copilot, Gemini, and how are students using them to cheat? Sure. James D. Walsh I mean, the vast majority of students I talked to were using ChatGPT, OpenAI's platform chatbot. And so much so that sometimes they would even refer to ChatGPT and then I'd go back to them and, you know, maybe they were using another platform, but ChatGPT was just kind of shorthand For a chatbot or AI. But of course, students talked about using Anthropics, Cloud and Gemini, Google and Copilot. And they were using it for kind of every facet of their education, really. You know, an English major and a writer where I was really interested in hearing how they were using it to write their essays or outline their essays or generate ideas for their essays. But students were also eager to talk about, you know, how they used it to take notes during class for them or summarize textbooks or create study guides and outlines. And then one of the biggest use cases is just computer science students who are using it to code. Some science students using it, you know, to analyze data. But the computer science students using it to code is just a huge use case. And I spoke with a professor, science professor at Berkeley, one of the top computer science schools in the country, who said his students have a serious problem. Tons of his students are using it on their take-home assignments. ([Time 0:08:48](https://share.snipd.com/snip/7d9e2ab2-7c43-4d35-aa8d-6d8acfb52d2e)) - **Trojan Prompts And Blind Submissions** - Students often paste professors' prompts into ChatGPT and turn in the output without reading it. - Professors embed weird tokens like "mention broccoli" to catch copy-paste submissions, yet many students still submit unchanged AI output. Transcript: Speaker 2 Yes. James D. Walsh I mean, it depends on kind of the type of student, the type of class, the type of school you're going to, you know, whether or not a student thinks they can get away with that is something Or whether or not a student can get away with that is something is a different question. But there are plenty of students who are taking their prompt from their professor, copying and pasting it into ChatGPT and saying, I need a four to five page essay and copying and pasting That essay without ever reading it. One of the funniest examples I came across is a number of professors are using the so-called Trojan horse method where they're dropping in kind of these non sequiturs into their prompts. Mention broccoli. Broccoli. Mention Dua Lipa. Say something about Finland. Into their prompt so that if you were to just copy and paste the prompt into ChatGPT, the essay that it spits out will say something about broccoli or Dua Lipa. And there are students who are just blindly copying and pasting those essays and handing them in. So not only, you know, are they not writing their papers, they're not even reading them. Speaker 2 Well, it seems like with just a little bit of effort, you could cover that up. I mean, unless you're just so lazy that you just insert the prompt and just copy and paste whatever you get back. But if you just take a little bit of time to comb through it, you should be able to cut that stuff out and cover up the trail. James D. Walsh You know, so many of my students are using AI, and I know that so many more students are using it, and I have no idea because it can essentially write, you know, 70% of your essay for you. And if you do that other 30% to cover your tracks and kind of make it your own, you know, it can write you a pretty good essay. ([Time 0:13:29](https://share.snipd.com/snip/244eb557-385f-4a2d-a29e-317dae9e5e23)) - **Student Who Treated College Transactionally** - Roy Lee used AI to cruise through Columbia assignments while focusing on networking and startups. - He treated college transactionally: find a co-founder and a spouse rather than deeply engage with coursework. Transcript: Speaker 2 Use it to hack. There's this guy, Roy Lee, who's an interesting character, sort of the protagonist of the piece in some ways. And he has this interesting path from community college to Columbia to startup founder. What did you make of him and his story. Sure. James D. Walsh Roy is somebody who just has always known he wants to be a founder, and he's been laser focused on that. And so, you know, he told me about going to Columbia and using AI to cruise his way through every assignment. He just did not care about the assignments. And because he had such a winding road to Columbia, I and his story? To stop him and say, like, why would you go through so much effort? You know, he took a gap year and then did a year at community college to get into an elite school like Columbia and then just not take advantage of it at all and not do the work. And he said, I'm here to find a co-founder and a wife. And I think he was really helpful because it's not just about his approach to AI. It's just about his mindset and his idea of college being this transactional place that, you know, gets you something very specific and is kind of the stepping stone. I thought that was really kind of instructive. Speaker 2 It does kind of give the game away a little bit, right? I mean, going to college, it's, you know, get out of here with this business about, you know, learning. It's just, it's a networking activity. Right. James D. Walsh And honestly, you know, as kind of funny or maximalist as his language, or like this example is, there's truth to it. ([Time 0:18:40](https://share.snipd.com/snip/94595ce6-cfff-4df4-8d39-9351fec54942)) - **Writing Practice Trains Thinking** - Using AI early can erode cognitive skills because offloading writing reduces the brain's practice at thinking through prose. - Walsh and some professors fear students will lose critical thinking and writing fluency by relying on AI. Transcript: James D. Walsh And I am somebody personally who, like, when ChatGPT, the version as we know it, in November 2022 came out, you know, shortly after was the first time I really played around with it. I was like, ah, it's a party trick. And I think in the course of reporting the story, I played with it a lot more to, you know, familiarize myself with it. And it was like the first time I realized, oh, if I, you know, had two paragraphs and needed a transition and I couldn't come up with it, you know, like it could offer me something. And I am, you know, much older than these students. And there was an immediate realization of like, if I start doing this now, I am going to lose something. Like there's some part of my brain is not going to, um, to, to flex work. And that is really scary. That's scary to me. And it's enough of a deterrent that like I just don't want to do that. ([Time 0:23:57](https://share.snipd.com/snip/ad23b5ec-a6af-4d47-872d-ee12202a0248)) - **Faculty Demoralization Over AI** - Many professors feel despair because AI use is widespread and hard to prove or police. - The volume of suspected AI-written work overwhelms instructors and undermines enforcement of academic standards. Transcript: Speaker 2 Let's talk about the professor perspective. I mean, the professors you spoke to all seem to share something pretty close to despair. Yes. James D. Walsh Those are primarily the professors in writing heavy classes or computer science classes. You know, there were professors who I spoke to who actually were really, you know, bullish on AI. I spoke to one professor, doesn't appear in the piece, but she is at UCLA. And she teaches, I believe, comparative literature and used AI to create her textbook, her entire textbook for this class this semester. And she says it's the best class she's ever had. And so I think there are some people who are optimistic. She was an outlier in terms of the professors I spoke to. For the most part, professors were, yes, in despair. They don't know how to police AI usage. And even when they know an essay is AI generated, the recourse there is really thorny. If you're going to accuse a student of using AI, there's no real good way to prove it. And students know this. So, they can always deny, deny, deny. And just the sheer volume of AI-generated essays or paragraphs is overwhelming. ([Time 0:31:21](https://share.snipd.com/snip/4c4c9a39-c7ae-497a-b178-8a6dd769a702)) - **Teach The Cost Of Outsourcing Work** - Emphasize to students why they shouldn't outsource work and what they lose by cheating. - Use education and conversation as primary tools rather than expecting policing to scale. Transcript: James D. Walsh Yeah, I don't, I somehow don't think we're going to police our way out of this problem. You know, one, also just like asking professors to like do this CSI sort of thing and then go to the, for every AI generated essay is just not sustainable. So, you know, what professors, you know, and administrators are kind of talking about is one, upfront getting students, the most important thing for them to learn at this point is like Why they shouldn't be using AI and convincing them not to use AI. You know, cheating in general is the kind of research on it shows that cheating comes, you know, from a lot of different factors that we kind of all understand. You cut corners when you're, you know, stressed or, you know, short on time, You know, there are all these external factors. And I think if we can really get students to understand what they're losing by cheating, that's like the number one kind of method of making sure they don't use AI. From there, like there's the practical stuff, like going back to blue books, going, you know, I talked to one professor who has switched entirely from essays to oral exams. And he really enjoys it. He sits down. I mean, he's lucky in that he's got small enough class sizes that he has time to do this. ([Time 0:34:12](https://share.snipd.com/snip/69834cbe-a598-4aa1-a5f5-8f2d9378eaf5)) - **Change Assessments To Verify Mastery** - Mix assessment formats: use in-person blue-book essays and oral exams where feasible to verify student mastery. - Combine these with teaching why students lose value when they outsource thinking to AI. Transcript: James D. Walsh So, you know, what professors, you know, and administrators are kind of talking about is one, upfront getting students, the most important thing for them to learn at this point is like Why they shouldn't be using AI and convincing them not to use AI. You know, cheating in general is the kind of research on it shows that cheating comes, you know, from a lot of different factors that we kind of all understand. You cut corners when you're, you know, stressed or, you know, short on time, You know, there are all these external factors. And I think if we can really get students to understand what they're losing by cheating, that's like the number one kind of method of making sure they don't use AI. From there, like there's the practical stuff, like going back to blue books, going, you know, I talked to one professor who has switched entirely from essays to oral exams. And he really enjoys it. He sits down. I mean, he's lucky in that he's got small enough class sizes that he has time to do this. But he has really enjoyed having this one-on time with students and getting to ask them questions and hearing their responses. And it's a better way for them to show them the mastery of the topic. But he also admits that there's something lost. ([Time 0:34:31](https://share.snipd.com/snip/020fbeab-44c6-4945-9554-dc04655fe75d)) - **Institutional Incentives Can Conflict** - Universities may adopt AI commercially even while failing to address classroom harms, creating a conflict of interest. - Walsh suggests admins sometimes favor adoption and revenue over confronting educational degradation. Transcript: James D. Walsh So they want to adopt the technology and aren't really picking up on how dangerous it might be. Speaker 2 I still know a lot of people who teach at universities, and I talk to them all the time. And a lot of them tell me that they feel very much on their own with this, that the administrators are pretty much just, hey, figure it out. And I think it's revealing that university admins were very quickly able to, say during COVID, implement drastic institutional changes to respond to that. But they're much more content to let the whole AI thing play out. And just so that it's clear what I'm saying, and this is me saying this, this is my opinion. I think that they were super responsive to COVID because it was a threat to the bottom line. They needed to keep the operation running. AI, on the other hand, doesn't threaten the bottom line in that way, or at least it doesn't yet. But AI is a massive, potentially extinction-level threat to the very idea of higher education. But they seem more comfortable with a degraded education as long as the tuition checks are still cashing. ([Time 0:40:38](https://share.snipd.com/snip/1735a062-8fd4-45cc-a4ed-6cb7551e44ec)) - **This Is More Than A Moral Panic** - This moment may not be just a moral panic because AI's power exceeds past tools like spellcheck or calculators. - Walsh notes that AI leaders warn of extreme risks, so treating it as trivial underestimates its transformative effect. Transcript: Speaker 2 People panicked about the internet. Hell, Socrates panicked about the written word. Um How do we know this isn't just another moral panic, one that might look silly in retrospect? James D. Walsh To be clear, this is my opinion. I genuinely, I don't know if it's the case. I think there's a lot of different ways we could respond to that. It's not a generational moral panic. This is a tool that's available and it's available to us just as it's available to students, society, and our do culture will decide what the morals are, and that is changing in the way That the definition of cheating is changing. So who knows? It might be a moral panic today, and it won't be in a year. However, I think somebody like Sam Altman, you know, the CEO of OpenAI is one of the people who said this is a calculator for words. And I just don't really understand how that is compatible with other statements he's made about AI, know, potentially being lights out for humanity or statements made by people at An anthropic about the power of AI to potentially be a catastrophic event for humans. So, and these are the people who are closest and thinking about it the most, of course. I have spoken to some people who say, you know, there is a possibility, and I think there are people who use AI who would back this up, that, you know, we've kind of maxed out the AI's potential To supplement essays or writing, that it might not get much better than it is now. And I think that would be a very long shot and one that I would not want to bank on. ([Time 0:51:00](https://share.snipd.com/snip/4f9e9c40-0871-4e02-94ff-fc87567142b8)) - **AI Could Widen Educational Inequality** - AI can both level the playing field and amplify inequality depending on access and pedagogical choices. - Walsh warns writing could become a privileged elective reserved for those with resources and time. Transcript: Speaker 2 Or do you think this might help in some way level the playing field? Is this something you thought much about? James D. Walsh Yeah, I mean, I thought about it a lot in the context of education. You know, there are certainly really good use cases for AI in leveling the playing field, right? As a writing tool, it can be extremely helpful for people writing it in their second or third language. It can be really helpful to create study guides for really dense material and, you know, put it in ways that are customized to your style of learning. That's really cool. You know, in terms of how it could accelerate these inequalities, that, you know, the idea that writing and thinking at a college level can be even more specialized, you know, I kind Of put a line in about it becoming, you know, that writing will be an anachronistic elective like basket weaving, right? It's like, it's going to be only for people who go to this certain school or can afford this certain school and have the time and privilege to write and therefore to, you know, engage with Ideas and think. And if that's the case, who's going to be able to afford to have those experiences? ([Time 0:56:06](https://share.snipd.com/snip/0b2ebfe0-64b5-48ec-a859-bcc0eb5532f2))