Clay Shirky leads AI efforts at New York University, and in recent months he’s had faculty approach him with a complaint: The advice he gave them on how to teach with AI wasn’t working. Meanwhile an anonymous NYU student posted a plea on Reddit saying he couldn’t stop himself from using AI on homework even though it was stopping him from learning. So Shirky is pivoting, looking for new ways to change assessments and encourage more social interactions among students.
Clay Shirky leads NYU’s AI efforts, and his background is as a media theorist commenting on how tech changes society. During the rise of social media 20 years ago, he was giving TED talks and writing popular books on how digital tech was altering our world.
To his surprise, though, he says that background in earlier tech revolutions hasn’t been much help in figuring out how college professors should adapt to AI. This tech is different, he says, and NYU is now pivoting from the early advice they offered on the technology.
Articles mentioned in this episode:
“AI Is a ‘Wicked Problem.’ That’s the University’s Specialty,” By Clay Shirky in The Chronicle of Higher Education.
“Students Are Skipping the Hardest Part of Growing Up,” by Clay Shirky in The New York Times.
“Can’t Stop Using ChatGPT on HW,” Reddit post, now deleted but on the Internet Archive.
“Guilty about using AI but don’t know how not to,” Reddit post.
“The Souls of the Machine: Clay Shirky says the Internet revolution has only just begun,” by Jeff Young in The Chronicle of Higher Education.
"AI Homework," by Ben Thompson.
“‘We don’t need to panic’: Clay Shirky on NYU’s approach to AI,” in Washington Square News.
“Inside the MIT Report on How AI Is Eroding Campus Culture,” Learning Curve Season 2 Ep. 1
“The AI Optimist Researching ‘Cognitive Surrender’,” Learning Curve Season 1 Ep. 21
“A Question of Signaling,” by Bryan Kaplan.
“The wicked problem of AI and assessment,” in Assessment & Evaluation in Education.
“Good Anxiety: Harnessing the Power of the Most Misunderstood Emotion,” by Wendy Suzuki.
Sponsor:
This episode is brought to you by Studiosity. Learning is a human act of thinking, not simply the file a student hands in. Studiosity champions the productive struggle. It happens in the moment when a student could easily outsource their work to AI, but decides to keep thinking instead. Studiosity provides the infrastructure to support that vital choice over and over. The work is supported, so students learn to think through it. Keep the learning real. Visit studioiosity.com/learningcurve. That's studioity.com/learningcurve.
Jeff Young:
Hello, and welcome to Learning Curve, where we look at how education is adapting to the rise of generative AI.
I'm Jeff Young, a journalist focused on the intersection of tech and education.
What if a lot of the early advice colleges gave professors about how to teach with AI was wrong? Maybe the guidance looked good based on past technologies, but doesn't actually fit the reality of generative AI.
We are four years out from when ChatGPT was first released, and these days I am seeing some examples of colleges shifting gears from their early reactions to AI.
And one college leader who doesn't shy away from admitting some pivoting is Clay Shirky, who leads AI efforts at NYU.
Clay Shirky has an unusual background for a college administrator in that he first made a name for himself as an influential pundit about how new media was impacting society. About 20 years ago, he was giving TED talks about how the internet and early social media was changing the world, and he was writing popular books, including "Here Comes Everybody: The power of organizing without organizations,” and “Cognitive Surplus. How technology makes consumers into collaborators.”
When I first met Clay Shirky back in those early internet days, he was a professor at NYU's Interactive Telecommunications Program, where he guided students who were building creative apps aimed at social activism or community engagement.
He has a background in theater, but taught himself to build web pages back when that was a very rare and in-demand skill. When I wrote a big profile of him back in 2010, I described him as a mix of dreamer, teacher, and commentator. These days, he wears a different hat as a high-level leader for NYU who has to talk to professors all over campus and think about tech policy and strategy for this large institution. His official title is Vice Provost of Artificial Intelligence and Technology in Education.
Clay Shirky:
And that big long title just means whenever anything digital hits the classroom, my job is to help faculty and students adapt. Right now, that's mostly AI.
Jeff Young:
I have talked to Clay Shirky many times over the years, and I've always appreciated his plain-spoken way of breaking down big ideas and seeing the larger context during times of rapid change. And lately, he's been writing some influential essays about how colleges should approach AI, and so I've been especially curious to talk through his ideas with him. In fact, I've been trying to get him on the podcast for several months now, and I ended up tracking him down at a conference in Denver the other day, where we finally set a time to connect. Okay, I saw you speak last week.
We were both at this Educause conference, the the big annual conference for college tech leaders, and I heard you say on stage there that you know as you help professors adapt to AI, that this job has been bigger a bigger lift than expected, and and you've been in this, you know, as we'll talk about as we go. Like you've been in tech for a long time. So what what why did this surprise you? Why has it been a bigger lift than expected?
Clay Shirky:
Well, it's been a bigger lift than expected for a bunch of reasons. I think for me, the deepest one is the problem I had that you've just alluded to. I've been in tech for a long time, I've been around for a bunch of these changes: the invention of the web, the appearance of smartphones, the rise of social media, the cloud. Because of my scholarly work, I've studied a number of other technical transitions: radio, television, the printing press.
And it took me almost two years to realize there's no good comparison. I kept trying to fit AI into previous stories. You know, people would reach for this is like GPS, this is like the calculator, and there are individual intuitions you can tune with things like that. But AI is hitting so many things at the same time, and it is simultaneous. Incredibly powerful without being incredibly good, that you know that you you have both this incredible leverage and these weird error cases.
When you think back in comparison with the rise of the spreadsheet, you know with the Bricklin and Frankston original Visicalc Spreadsheet. It only did one thing, but it did it really well, and it got it got good before it got powerful.
AI got powerful before it got good.
And so it's been disorienting to me to say you know I'm old. I've been doing this a long time. I have this wealth of stories, and those are actually hampering my ability to write what's right in front of my face.
And so I spent too long, frankly, thinking we can design our way around this. Right? Faculty need to keep students doing their own work plan to the point of being in a classroom for whatever the subject is to get the students to learn by doing. All learning is learning by doing, in fact.
And we spent too long at NYU thinking we can redesign the assignments.
And there was a moment; it wasn't even my revelation. I went and was speaking to a group of faculty, and there was a philosophy professor who's actually become a friend because he was he was so whenever this this conversation was so direct he was not happy to see me coming at his arms crossed across his chest like literally arms crossed across his chest like who's this guy from New York coming to coming to lectures I was in in our Abu Dhabi campus at the time and he said in in blunt terms we tried what you suggested and it didn't work. Is that the students?
You know, you can talk to them as movingly as you like about the need for them to engage and do their own work and their own learning. They are opting out, or enough of them are opting out in the classroom that we can't manage the classroom the way you're suggesting.
Jeff Young:
And your suggestion had been basically what?
Clay Shirky:
The suggestion had been redesign the assignment so that AI is part of it, but that there's a human element. All that stuff we used to say: multimodal, personal engagement, focus more on process than product.
And those weren't bad ideas, but they weren't effective. And then the other piece of it was to talk to students about the sort of shared goal, like everybody's in this class to make you smarter, right? That's the goal of the whole enterprise, and students understand that intellectually.
Everyone knows that if you rely too much on the tools, you're offloading your own thinking.
But man, if it is 11:30 at night and you have a physics problem set and a 1500 word essay, do both in the morning. You know where the shortcut is going to appear.
And so what Maddie, my philosopher friend, said was, "This is not good enough. And at that point, I began to look around and realize we have to redesign assessments. We can't just redesign assignments. And the terrible thing about redesigning assessments is we brought students back to oral exams, blue books, Socratic dialog, you know, office hours, and faculty and students report enjoying these things. Right, there's more direct engagement with the professor, but it is we are also risking benefiting students who think quickly rather than students who think deeply, and that shift to extemporaneous short-term production. Here is a blue book. Write an essay. I am a terrible writer. My first drafts are awful. No, seriously, like
Jeff Young:
That's interesting because you've written so many pieces that you know that books.
Clay Shirky:
I've I've written a bunch, but I whatever somehow I've contradicted the 10,000 hour rule by not getting any better at it.
Like I just I read my own first drafts. Like oh god, half of this has to go.
And with blue books. You do not get that. Yeah, do not get that with books.
So we're in the middle of this this transition, which is as I said harder than I expected, precisely because Because the obvious balance we could all articulate in 2023, there are good uses of AI and bad uses of AI, and we want to, you know, move the students towards the good uses and away from the bad uses. I spent about a year and a half thinking that is one single problem. Right?
There's a spectrum, and we want to move students to the good end of the spectrum. Easy enough, but I longly assumed that moving them towards the good uses would also move them away from the bad uses.
Jeff Young:
I see. Yeah.
Clay Shirky:
And it's not a spectrum. We have two different problems, which is to say, a student can absolutely understand. I'm using this thing to spar with. I'm having a dialogue that changes my ideas. I've got a Google notebook that is you know helping me you know extract things from from academic resources.
All that stuff can be good. It doesn't take the ability to shortcut out of people's hands.
The research that came up around this. I think it really started with the conmigo with Sal Khan's math tutoring chatbot, where the headline was: "Students who use these tools get a lot better at math. Students who use these tools as directed get a lot better at math. The footnote was only one in 20 students use these tools as directed, and so that has proven to be harder than the web, harder than the arrival of mobile phones in terms of change to campus culture.
It's a really hard problem, and I think again for people who've been around a while and who pride themselves on knowing a lot of stories about various different tech transitions, the ways that this one is different. I think we were people, you know, basically sort of my age were slow to recognize because we kept trying to analogize it to something else, and it took a while to realize, no, no, this has got a bunch of characteristics which, taken together, are not really like any other digital transition,
Jeff Young:
Yeah, nd that's that's so interesting about, and it also, I I think I can relate just personally having covered those other trends that we that you mentioned on the internet and the and going through that and yeah I hear you this is this is different and so it's so interesting to hear your experience there on campus.
So, okay, you wrote a piece for the Chronicle of Higher Education in August, so not long ago, and it got a lot of buzz. I heard a lot of people reference it as I was doing my work, and it you called it the headline is AI is a wicked problem. That's the university's specialty. So now there are a lot of campuses. There are a few anyway that are getting a big a lot of press, like places like Purdue and Ohio State that are saying like we're you know at Purdue they're saying they're going to change all the majors to incorporate AI fluent fluency, change the whole curriculum, to have like one huge you know kind of it university wide effort, but you're essentially arguing that the best answer for a college on AI is to be incoherent. It sounds like
Clay Shirky:
I mean I guess I'd make an even deeper argument, which is incoherence is our stock in trade. Okay, that what do
Jeff Young:
You mean by that?
Clay Shirky:
So we host individual faculty members, and sometimes even whole departments that have incompatible worldviews. Right? We do like we we don't have a divinity school, but the universities that have divinity schools do not make the priests in training and the physicists in training sit down and hash out what the really actually true view of the universe is.
Right at NYU, we hosted something famously called the Sokol hoax, where one of our physicists wrote a bullshit article and sent it to a journal of theory, which was then. Oh,
Jeff Young:
this is all pre AI, right? This is all pre AI fabricated research.
Clay Shirky:
Late ‘80s, early ‘90s.
Jeff Young:
That's right. So big scandals
Clay Shirky:
Yeah, both of the professors involved worked at NYU. Both the physicist and the publisher of the post, the journal of essentially postmodern theory.
No one ever said, "Okay, the two of you need to sit down and hash this out. Right. So what universities have always done is host long lasting, and I mean, in some cases, centuries-long conversations. The famous struggle between the ancients and the moderns was the modern world something that we've fallen from and can only attempt to retain that glory, or are we headed into a bright, beautiful modern future where we're building on them? That fight took centuries, and we hosted it the whole time. So the idea that the university would be better off if we were more uniform, more rational, more coordinated, just seems to me to be out of sync with the actual story of the last 1000 years.
Jeff Young:
So in other words, what you're referring to here is that you know scholars in one department might not agree on the direction of their discipline. So why should everyone on a campus have the same AI policy?
Clay Shirky:
Exactly, exactly. Look, the chemists and the choreographers are using these tools differently right now. I mean, anybody who wanders into two different classrooms is going to see a different faculty approach, right?
Again, going back to this this fateful conversation with Matty Silverstein with with my my philosopher friend, he one of the things he said in that that first meeting we had is I keep hearing that it's okay to brainstorm with AI as long as you're writing, he said, "I teach a philosophy class. Like I'm trying to get these students to have an idea and then put it under enough pressure that it becomes worth talking about. Like brainstorming is the last step I want them to skip.
So it's like he'd almost ….
Jeff Young:
You mean rather them have it write the paper once they come up with the idea?
Clay Shirky:
Yeah, right.
Maddie doesn't want AI at all, but he was especially offended by the sort of general consensus that the further upstream you went, the better off you were letting the students use AI.
Can I? Can I? Yeah,
Jeff Young:
Yeah. I guess for me, the interesting thing on that, which I think has gotten a lot of fans, this argument of reminding the history of the university is to have all this, you know, differing opinions.
Now, from a student perspective, that could be in this case when we're talking about teaching approaches. It seems like it can be frustrating, right? They're hugely annoyed. The students are
Clay Shirky:
Hugely annoyed by because
Jeff Young:
They don't know what it is that that you know is it the devil or the thing they should need to know? Otherwise, they're not going to get a job. Is it going to save them or kill destroy their learning?
Clay Shirky:
Yeah, and there are I think two different problems there. Masquerading is one.
The internal problem is as between faculty approaches and departmental approaches. The external problem is the literally millennium old tension between higher education as a proving ground for individual self-development and higher education as a job training program. There is a group of faculty who believe that you know some somehow up until the late 1950s students only came to college for the love of learning and you know all the all the employment stuff was downstream. It's ridiculous.
The first university was a law school. Like anybody to present exposure to literature as the kind of core function of a college, just doesn't know anything about the first 300 years of our history. Right? It was it was law, medicine, and theology with the first three faculties and the first degrees were master's degrees, not bachelor's degrees.
The whole edifice of humanism comes in hundreds of years after we get started. So anyone talking about the essence of college education is really just cherry picking from this crazy 1000-year history where there's not a lot of whatever. There's not a lot of sort of uniform lessons to be drawn.
So the internal question: What we've done at NYU is we've said to faculty, you don't owe students consistency, but you do owe them clarity. That and the students, you know, my colleague Deangelo and I every semester will sit down with a group of students, you know, 10 or 12 students. Chatham House rule: We're not going to quote you. Just tell us what's going on with AI. Nobody's tracking you.
Nobody's going to, you know, nobody's going to go after figuring out what classes you're enrolled in. They're pretty frank with us. I'm sure they they pull some stuff back, but in those conversations, they often independently come to the realization that they don't want their professors to all have a uniform policy, because they start to understand themselves as like, oh, I'm I'm majoring in physics, and my friend is majoring in biology or in comparative literature. Yeah, like you know, and the scholarship is starting to build up around this.
The second thing, though, Jeff, the anxiety about the job training is the harder one in a way because universities have this kind of relationship to job training. Yeah,
Jeff Young:
Yeah, you're putting your hands like they're both like loving and hating them.
Clay Shirky:
Push back, pull it forward at the same time. We love to take credit for our students getting good jobs after they graduate.
Clay Shirky:
And at the same time, we don't love the idea of trimming our curriculum to short-term business needs,
Jeff Young:
Yeah.
Clay Shirky:
And I will say, Angie Comith, our wonderful dean of the School of Professional Studies, so our school that's most involved in direct conversations with employers, has said every time we go to employers and we ask them to help us design the curriculum, like let's do this together. We all want our students to get jobs. The employers always try to offload the cost of the first six weeks of training, right? They're not actually thinking long term, and yet when we ask them, "What do you want in a graduate? like we want critical thinkers, we want good collaborators. So we're the ones who have to sort of hold back from the immediate job training thing and say, "Look, you want flexible people who long term can adapt. You don't want them spinning up on today's technology. So, look, different universities have different missions. If a community college says we're graduating people who are going to go work with exactly these tools in exactly this situation, right? Every machinist who works at Boeing has an associate's degree in advanced, you know, advanced production. That is their function.
We're not that. We're a private research university, and so we're more interested in figuring out what are the long term characteristics of a human being for operating in an AI world. You may remember the head fake of prompt engineering a few years ago, where everybody was going to learn prompt engineering, and of course….
Jeff Young:
There were classes in it. Yeah, at colleges, I am proud to say
Clay Shirky:
At NYU never once went in for that bullshit.
We looked at that, we said the tools are going to get good enough that this whole thing is going to go. Reasoning models basically destroyed prompt engineering because they were. Better at taking apart an initial question than you could you could train a prompt engineer.
So I think a lot of these press releases of the sort that you announced they do signal like this university is committed to these students being able to graduate and use these tools, but they're doing it at a time when the businesses themselves often can't articulate what they want from AI, and so it it may be useful as a signaling. You know, Brian Kaplan's work on college is a signaling effect. He's not wrong about that. I think he's too dour about it, but it does.
We absolutely do create that value in the economy.
But I'm skeptical that the curricula themselves that are being designed around this idea of AI fluency or AI literacy are going to survive five years of development of AI itself. And I think there's never been a time when you ask businesses what they want that they don't say we want critical thinkers, we want good communicators, we want good collaborators, we want people who can make decisions under pressure with imperfect information.
None of those are AI things. You can use AI in any of those skills, but I'm. I think the short term shift to workplace readiness and away from the kinds of things people have historically wanted from college graduates. I don't think it's going to last.
Jeff Young:
That's yeah. I there's so much we could dig into, but I have some other things I wanted to get to.
In an article that you wrote for The New York Times earlier this year, you described a situation at another university. It wasn't NYU, but where students got caught cheating with AI, and then as part of their punishment, they had to write an apology letter.
Clay Shirky:
Yep.
Jeff Young:
And some of these students used AI clearly to write these apology letters. They didn't do it. So, what do you make of all this? I mean, this is just one example. You, you, and others have cited plenty more, where just this wow, this like you're just like okay, students did that with AI and that and that, where they kind of knew it wasn't, you know, was not allowed.
Do students today have a different sense of ethics, or like, is there something kind of what is going on?
Clay Shirky:
Yeah, I mean, look, if the easiest thing in the world for somebody my age is to be like, kids today don't, you know, nobody gave us the chance to do those things in college. I cannot say actually that we would not have done them. So I certainly do not want to get on that particular high horse. Sure.
What I will say is behavior is motivation filtered through opportunity. AI has radically transformed the opportunity landscape.
The idea of getting out of the emotionally difficult work of apologizing to someone was never really an option before. You go to your friends and be like, "I got to write this thing to my professor. What should I say? But you still, you were going to do the work, and frankly, even conversations with your friends was emotionally engaging in a way that was, you know, whatever tedious and painful or involved some self reflection, which is never fun.
Once that became possible, people who have identical motivations to anybody in an older generation nevertheless have a different opportunity landscape. But it is also the case, and I think a bunch of you see this in the recent MIT report on sort of yeah…
Jeff Young:
Which we did an episode on.
Clay Shirky:
Oh, yeah, the principles which I love that report the principles approach rather than the policy approach, they said basically these students are under so much extrinsic pressure for grades in order to get jobs and be considered for you know the kinds of companies they want to apply to.
You know, when I was in college, we didn't worry about employment as much as I see our undergraduates worrying now, because the economy was more beneficent and because there were fewer colleges. We were just in a different competitive landscape.
And again, it's very easy to say, "Oh, well, we were ethical, you know, way back in the late 1930 s or whenever we were in college, whereas whereas these kids.
Jeff Young:
you really put us in a black and white movie all the time.
Clay Shirky:
Oh my god, my whole life now. When I talk to young people, I see my past in sepia.
Jeff Young:
Right, right.
Clay Shirky:
But it would be very easy to say, well, you know, in my day, we didn't worry about this. We didn't worry about it because it was obviously going to happen. Everyone assumed we would get a good job. We didn't even think about it.
Jeff Young:
Sure.
Clay Shirky:
It was just if you got this degree and you went into that workforce, you got employed and you got a good job. And now that sorting out is so internal to the college experience.
The biggest single mistake we have made as an industry in the last generation was to start sharing GPA and transcripts with outsiders. Those used to be inside. That used to be inside only data. Right. The only thing you would know about a graduate of you know Ball State or whatever was this person did or did not graduate. Graduated with this major, and they did or did not.
Jeff Young:
Why is it bad to share that?
Clay Shirky:
Because of Goodhart's law, a metric that becomes a target ceases being a good metric. All of the grade inflation, all of the grade grubbing, all of that distortion happens because faculty know that the grade is not just feedback to the students or to other insiders.
I'm saying to any place this person might apply to graduate school, this is an A student or a B student. All of a sudden, whether or not you get a job at Accenture depends on what grade I give your sophomore year. Like that, the anxiety from the future has been pulled into the present and shot through every place where we assess our students and the faculty.
You know, grade inflation is presented in the outside world as being this obviously ethical lapse on the part of the faculty. It is, in fact, by and large, a sense that faculty are not going to handicap our own students right into the competition.
Jeff Young:
It's almost like the relationship the professor has with the student is sort of keeping them from being. I think. I think I saw you say some. They're both the referee and the the coach. That's exactly. And those are two different roles,
Clay Shirky:
right? And we can keep those roles going if we're in a place where I can sit down with you as a student and give you honest feedback. But I can also say to somebody else, this student passed my class, and they should go on and do X, right? All you need to know is, did they pass the class? Right. The distinction between getting a B plus and an A minus should not be meaningful to the outside world. And it was a mistake to ever let the outside world look at those metrics because they've been squashed flat.
Jeff Young:
Is there anything you could do about this at this stage? Like, sure,
Clay Shirky:
You can stop grading. Right. You can say basically pass.
Jeff Young:
And there is this movement to, and in fact, there we're seeing examples of universities try to remove grades from early classes. Yeah,
Clay Shirky:
And the other radical solution is to publish the students' grades and to publish the grade distribution for the class, right? Because a big part of like everybody gets an A is the outside world can't see into the institution's behavior as a whole.
You published a curve that said this person got an A and this many people got an A. You would be in a way reducing grades, the salience of grades for the outside world. It would be another way to attack the Goodhart's law problem of metrics becoming targets. But there is no way to force faculty to be cruel to their students.
You just can't do it, and …
Jeff Young:
That's maybe a good thing. But it's these structural changes are tough, and I liked the MIT report as well. Like people should go back and listen to this episode because it does talk about this holistic issue, and it talks about one of its recommendations is grades ought to be revamped.
So it's so many things. This is the wicked problem, I suppose. This is what you mean.
Clay Shirky:
Well, this is.
I mean, you know, this is the Deakin University paper, which made just a huge impression on me last year in 2025, and it starts with assessment, AI, and assessment. The how do you know that a student's done their own work? And it says basically this is one of these classes of so-called wicked problems.
These are problems where you have to act. There's no way to test a solution. Every solution is sort of at scale in one shot. There's no clean definition of the solution. If you change the people defining the problem, you change the nature of the problem. There's no agreement on what even needs to be solved, so you can't you can't start with consensus on what a good output looks like. And many policy problems.
The original wicked problems paper said basically, at a time of enormous hope for technocratic scientific management of society. What the original paper said is many policy problems are not solvable in scientific terms because they're not well defined in terms of either problem or solution or even stopping rule. Like we have solved this problem now. Greeds are one of those things. Assessment is one of those things. And what AI has done is it's taken a bunch of uneasy truces we had around a whole bunch of these issues and returned them to us in their unsettled form. Right? It's not like great inflation wasn't something we had to deal with before
Jeff Young:
it would come up. There'd be articles in the mainstream press every couple years, right? Exactly
Clay Shirky:
Right, exactly right. But a bunch of the trade-offs that we had kind of achieved some sort of equilibrium stayed on that was never perfect. You couldn't have perfect for a lot of these problems, but was at least broadly supported. AI has put pressure on so many of those things at the same time that you will periodically see these calls for we need a rapid wholesale reinvention of higher education, which I'm obviously skeptical of. We need to do the thing we've always done, which is we need to adapt slowly and incrementally, and each adaptation affects each other. Adaptation, which is the nature of the wicked problem, higher education does better when we adapt slowly but constantly, rather than when we present ourselves as being in pure stasis and then suddenly needing radical reinvention.
Jeff Young:
After the break, how a plea for help from an NYU student on Reddit changed Clay Shirky's thinking on how to respond to AI.
Stay with us.
Sponsor:
This episode is brought to you by Studiosity.
Policy dictates one reality, but the campus lives another. Studiosity provides the essential infrastructure to defensively validate the genuine learning journey, moving the sector towards a model of support and validate by replacing vague AI scores with binary clarity, Studiosity empowers students to stand up for their own work and self-validate against a personal baseline. This builds persistent authorship continuity, shifting students from passive suspects into active owners of their own thinking.
Simultaneously, Studiosity enables educators to set clear expectations using the AI assessment scale, giving them back control over assessment security and replacing detective fatigue with a richer story of the cognitive learning process.
After all, learning is a human act of thinking.
Keep the learning real. Visit studiosity.com/learningcurve/ for more details.
Jeff Young:
I want to plug another project that I'm working on these days. It is a newsletter about AI on campuses that I'm writing for the Chronicle of Higher Education.
It's called Jagged Intelligence. That name is for a term used to capture how AI is both surprisingly effective at some things and really bad at others, like those hallucinations it's prone to. I hope you'll check out the newsletter and participate in our regular flash polls to share your hopes and concerns about AI.
Sign up at Chronicle.com/newsletters.
Now back to the episode. I listened to a podcast you did. It was about a year ago, and one of the things that came up in that interview was a Reddit post that you said at the time was going around, and the administrators were particularly looking at it. It was by an NYU student posting on Reddit about their struggle with AI use. Could you? Could you? You're nodding. So could you remind what was this student's reaction, and and why did that strike you?
Clay Shirky:
So it was really striking because it came at around the same time of this massive reassessment of we can't just redesign the assignments and and and sort of make make our peace with AI in the classroom. The title of the post was "Help! I Can't Stop Using Chat GPT.”
Jeff Young:
An anonymous student whose student?
Clay Shirky:
It was, but a student at our engineering school. We don't know what department they were in, but the rest of their posting history was consistent with them being an undergraduate at NYU, as reported.
And they said, "I'm using this for everything now, I sort of ask it about my homework all the time. I can feel myself learning less, but I can't stop. And it was that sense of helplessness, almost at a level of addiction, a phenomenon we later went on to call cognitive surrender, but we didn't have that phrase at the time.
What that student was describing, I would later describe as cognitive surrender, which is I got so accustomed to this tool that I use it even when I know it's interfering with learning. I use it even when I know that I know the answer. And anybody who's ever typed seven times eight into a calculator knows this effect, right? Of I actually memorized the times table. I've known it for many decades now, and yet if the calculator's here, I'm just like seven times 15, whatever. I'm just not even going to bother.
Jeff Young:
I want to stop for a minute and dwell on this student post on Reddit that we're talking about. Again, the title of the post was "Can't Stop Using ChatGPT on homework.
I want to read this whole post because it really does seem like this cry for help.
The NYU student wrote, "I literally can't go 10 seconds without using chat when I'm doing my assignments. I hate what I've become because I know I am learning nothing, but I am too far behind now to get by without using it. I need help. My motivation is gone. I'm a senior, and I am going to graduate with no retained knowledge from my major.
I'm guessing some listeners might hear that student as lazy or. Be just kind of really unusual, but I am seeing these kinds of posts by college students and high school students around the country.
This is a trend.
Here is another one, a more recent Reddit post that was titled "Guilty about using AI, but don't know how not to. Hi guys, I've been feeling extremely bad about using AI because I know it takes away from the learning experience and damages the environment. This one is a little longer, so I won't read the whole thing, but I will put the link in the show notes. Anyway, it goes on to say, I try my best to always do my work by myself before I turn to AI, but it makes me realize how trash my original work is, and it makes me want to rely on AI more. It's so bad. I genuinely don't know what to do. I know I probably don't deserve to be in college with my capabilities and learning speed, but what can I do? I still need a degree. Do you guys have tips? I just want to feel proud of myself again, but I don't feel like I have the time to solely rely on my slow learning. I reached out to several of these students on Reddit, asking them if they'd come on the podcast and talk about their experiences, and I said they could stay anonymous. But so far, no takers. I am going to keep trying to find a student to talk about this phenomenon on a future episode,
Clay Shirky:
And I'll tell you when you were asking what I'm worried about. I'm always behind I'm always behind the curve, unfortunately.
But there was the original worry about cognitive offloading, and this is, you know, the conversation about not just assignment redesign, it's assessment redesign.
Then there's cognitive surrender, which is oh this is a holistic problem. This is not just each individual student can opt out of each assignment, but a student can actually hand over whole parts of their educational development to the tool.
And the two, the two findings of cognitive surrender that were alarming.
The first is once you get used to using it, you use it even when you know it's dumb to use it, and you just can't help yourself. But the other work that's coming out recently is when you give students these tools, the quality of the output improves exactly what you'd expect.
When you take the tool away, the quality of the output goes down exactly what you'd expect, but it goes down to below the baseline. It goes down to so it isn't just that when you take the tool away, you remove the artificial advantage.
When you take the tool away, it turns out that the student's mind has weakened in the period that they were using the tool. Yeah.
So it isn't even just like, well, I can you know I can wear these shoes and run faster. I can wear my regular shoes and walk at my regular speed.
It's as if running shoes make you walk more slowly when you're not wearing them.
And that worry came along, and most recently the worry has been around emotional offloading. So you alluded to these students who wrote their apology to a professor for using Chat GPT with Chat GPT.
Jeff Young:
Right?
Clay Shirky:
Some of that is time saving, but it doesn't take that long to say yes. You caught me. I'm sorry. This is dumb. I won't do it again. It's really opting out of an emotional connection with another person, and we are seeing more of that at NYU.
I talked about my colleague D'Angelo and I having this you know biannual conversation with students just tell us what's going on. And a year ago, I guess it was a year ago, a little over, students started talking about being in group projects and their fellow students having conversations with them using AI.
They weren't even opting into an ungraded, unobserved human conversation between students.
Jeff Young:
Wow!
So even in a group setting where you're just like they were sort of phoning in their group participation with ChatGPT.
Clay Shirky:
Exactly, exactly.
And that once I mean when students start opting out of conversations with faculty, we're authority figures. We're annoying them. Like I get it, right?
I've been a student. I've been a faculty member. I've been on both sides of that conversation. I understand how that can happen when students start opting out of conversation with each other, and and part of what it's a it's a bigger deal, and part of what worries me about that is that we tell the story of students' intellectual development and of faculty teaching students, but the truth of the matter is, in a full course load, we have students in our classrooms about 50 working days a year.
That's the what the number. That's what the hour.
Jeff Young:
Yeah, it's like a pretty small number.
Clay Shirky:
It's a pretty small.
Jeff Young:
It's like the number of minutes NFL players are like on the field, doing something. It's like equivalent, like roughly this kind of idea.
Clay Shirky:
Yeah, and so the and if people look back on their own college experience, so much of college is peer engagement, peer pressure, deciding who is admirable, who you want to be like, who. You don't want to be like if social, yeah.
And if students opt out of intellectual and emotional connection with each other, we could change nothing. We could keep our classrooms going identically to the way they were, and the quality of the entire educational experience will suffer.
And so the thing I worry about most now is if we're given the shortcut to having emotionally engaged conversations, and particularly if the students are given the shortcut to having emotionally engaged conversations with each other, that threatens the fabric of the enterprise that isn't any longer. Oh my God! A student passed their biology homework by you know figuring out the difference between or rather getting the difference between meiosis and mitosis from a from a chatbot. That is really college stops being a site of social transformation, right?
And this is one of the things that people get wrong about us in the sort of Silicon Valley rhetoric all the time. We are presented as being a hugely inefficient source of information transfer, which we are. That is, we do that, but we do it inefficiently.
But that's because we're not mostly in the information transfer business. We're mostly in the social formation business and people don't come to college to learn new things. You can do that from YouTube. You can do that from Wikipedia. They come to college to be different people, right? They expect to graduate with something more than just I learned 200 hours worth of fact or 200 working days worth of facts, right?
And the people who want us to just be job training systematically underestimate the social value of being part of this community for the time you're here, and if the community goes away and everything else about the institution stays as good as it was, we still all suffer.
Jeff Young:
I want to come back to what to do about that in a minute, but I first want to get at something that a lot of yeah right.
But I first I did want to ask NYU like most call like the majority of of higher ed institutions these days. By the way, I'm not trying to call you out really. Gives away free access to an AI tool. I think it's Gemini on your campus. Yeah, we're
Clay Shirky:
Google Shop so Gemini. Yeah.
Jeff Young:
So now, I mean, how do you square that?
Clay Shirky:
Yeah, yeah, yeah.
Jeff Young:
When these students are telling you, "Please help me get off this tool.”
Clay Shirky:
Yeah, yeah.
So that's an absolutely fair question. I've had to say that to faculty who have the same same question. Yeah. No,
Jeff Young:
I'm sure you're hearing this, right?
Clay Shirky:
There are two different answers: one for faculty, one for students. The answer for faculty, funnily enough, is academic freedom. If you went to a faculty who refuses to use AI and said, "You know what?
We held a vote, and most faculty are using it, so now you were forced to use it in your classroom. Uproar! That press release would not last 10 minutes before we had to reverse it. But the reverse is also true. The faculty who have thoughtful and deeply committed ideas about avoiding use of the tool cannot force the faculty who are using the tool to stop. So this is poorly appreciated because central administration gets shit for this at many many institutions of higher education. But outside of a really narrow set of ethical, legal, or contractual concerns, we cannot force faculty to do things if there's disagreement among the faculty themselves. So, on the on the faculty side, the reason we provision a tool is that faculty have demanded one. We were getting calls from 2023, like you have to license a tool, you have to license a tool. For a while, FERPA compliance was an issue.
None of the Frontier Labs wanted to deal with FERPA compliance and all those kind of technical things.
Jeff Young:
These are the federal privacy laws that protect student data and their their grades and all this. Right, yeah, right.
Clay Shirky:
And and and we just weren't going to sign a non-FERPA compliant contract with any of the Frontier Labs. Google finally came through, but it also means I have to go to faculty who have really strong ethical objections to AI in any dimension and say, I am not, and central administration is not a vehicle for you to force faculty you disagree with to stop doing what you dislike.
Jeff Young:
I see.
Clay Shirky:
This is, and this gets back to the incoherence conversation.
There's going to be a long, long conversation on campus about how AI fits into what we do.
But when some faculty demand use of the tool and other faculties demand that we not use the tool, we are a little stuck. There is no non. There is there. There's no solution that doesn't mean that people at both ends of the spectrum don't get what they want, and I will say I spent most of 2024. Anytime I met with faculty, somebody was going to be mad at me, but I never knew why. So sometimes
Jeff Young:
Or which way toward yeah. Sometimes
Clay Shirky:
I would go into a meeting and people would be like. Like this is the most incredible tool that's been invented in the last 100 years. It's basically child abuse not to give this tool to students right away. You are slowing things down. Why are you doing this?
And then faculty saying this is terrible. This is terrible for intellectual development. This is interfering with my classroom. Why are you allowing students to use this? So on the faculty side, we are effectively rendered unable to make an ethical call precisely because of the disagreement among the faculty. Mostly, what we try to do is host that conversation rather than settle it.
For students, the problem is very different.
No one uses Gemini. They don't trust us.
They use ChatGPT, or sometimes now they use Claude if they're particularly Claude Code. Students do not believe us when we say no one is tracking your information. Google is not training on your information, right?
Jeff Young:
Which is the reason you bought the tool. Like they're paying for a license so that they would use this. It's just as good and better what they've got for free. And we would like them
Clay Shirky:
to use the tool that has better protections that we have negotiated on their behalf. But they have consistently told us they do not trust us, and so
Jeff Young:
And you're probably seeing that in usage too. They're just absolutely
Clay Shirky:
No we go to the usage and it's tiny, tiny, tiny. We have I mean we're just looking at one particular tool that might be used. We have a campus of 60,000 people. 10 students had used it.
Not you know, if it was if it was 100 or 1000, even 10 out of 60,000. It's not even a rounding error.
Jeff Young:
but yeah, you're paying for it as an institution.
Clay Shirky:
And yes, we're paying for it as an institution. And so, look, it's a bundle, and so we're not so much looking at headcount and cost as we are thinking.
You know, it's a little bit like a teenage drinking thing. If students are going to be doing this, or you know, if your kids are going to be doing this, you'd rather be them doing it in your home, right?
We'd rather students be using it in a place where we know they're not leaking their data. We know that the you know that their data is not being absconded with or or trained on or whatever the things people are concerned about are, but those people raise those worries in the abstract, but the very specific worry that students do not want us to see how they're using the tool.
There's a deep sense that they're getting away with something. Even frankly, if they're using it in mostly approved ways, they just don't want to be thinking about it.
And I remember my colleague Deanzo put together a sort of student-to-student guide about AI, separate from what we did. And I remember going to address that group and talking about students using this work to opt out of the assignment, students cheating, and I called this lazy. I said basically students who don't do their own work are lazy, and the students couldn't even bring themselves to use that word, much less the word cheating.
They were not willing to pass judgment on their peers for using these tools because it was such a godsend in terms of reducing anxiety, and particularly, you know, as you remember, Jeff, a couple years earlier, post COVID, reducing anxiety has become this imperative.
Jeff Young:
Yeah.
Clay Shirky:
Whatever you do, don't raise the anxiety of your students. And my colleague Wendy Suzuki has written a whole book about positive anxiety-the anxiety of the jet pilot before takeoff, the actor before they go on stage.
Jeff Young:
Well, the sort of the driving anxieties that are positive in the way. Right.
Clay Shirky:
So we're trying to have a conversation with students about anxiety as a focusing force rather than as a debilitating force.
Jeff Young:
Sure.
Clay Shirky:
That conversation takes a while, and in the meantime, are. Offering a student a tool does not take other tools out of their hands. So when faculty say you are giving these tools to students, that is true in terms of the contract. It's not true in practice. In practice, the students are going to tools that they are either using the free version of, or are even sometimes paying for themselves, but that we don't have our fingerprints are not on it at all, and that is a cultural change that's also going to take a while to to shift.
Jeff Young:
This really is worth highlighting because it's kind of a crazy moment that we're in with AI in education. Most colleges around the country are paying for licenses for ChatGPT or Gemini or Clod for their students, even while many students are refusing to use those college-provided accounts and opting for free versions instead. I'm going to be looking into that trend more as well. But as I talked to Clay Shirky, I was most interested in how the presence of AI is changing social interactions on campuses. I think this is a really big deal, and it's turning out to be a key theme of this season of learning curve. And the social part.
Let's get back to that for a minute.
Do you have any thoughts or plans or initiatives that you're considering about how to address that?
Clay Shirky:
We do, and it's one that I. Actually, inherited from a colleague of mine, Christy Patton, who's now the dean of our education school, was a colleague in central administration when she started this. She started an initiative called IRL in Real Life, and it was originally just meant to address the creation of social opportunity for students at NYU. Now, our campus, anyone who's been here knows that our campus isn't really a campus; it's just New York City. It's like, yep. You walk out of a building and you're at a subway stop, and the closest thing we have to a central campus is actually a city park, Washington Square Park.
Jeff Young:
Yep, I love that neighborhood. The fountain. It's a storied place. You've seen it in movies a few. Yeah, indeed.
Clay Shirky:
Yeah, wandering around, wandering around Greenwich Village. I live on the street where the Bob Dylan Bleecker Street photo was taken. There you go.
But students have consistently complained, like it's hard to find my people. It's hard to find community. There isn't really a commons. There isn't a quad. There isn't that sense that you know some both smaller and less urban universities have.
So we've always struggled to create kind of communal opportunities for our students because we want everybody who wants that experience to to be able to have it.
Jeff Young:
Sure.
Clay Shirky:
But then mobile phones, social media are adding a further mask.
So Christy starts this thing called IRL, which is just we're going to start hosting parties where there's a phone check-in by the door. Everyone checks in their phone, so it's not like you personally have to give up your phone.
It's when you walk; it's you know like speed dating. Everybody's doing the same thing, so it feels less awkward.
You walk through this door; no one has their phone in that room. Anybody can walk out and get their phone anytime they want, but people report having a much better time, right? And so we start doing all of these IRL events and faculty say, "I want that. I want that for my classroom. I want sticks on my classroom.
So more and more faculty are going to device-free classrooms. Yep. Interestingly, I just heard from our law school who started out saying we are taking the opposite tack of much of the rest of the university.
We are starting with the assumption that these tools exist, and we're asking ourselves, why wouldn't you use them?
They still ended up with device-free classrooms. They still ended up with in-class writing. Like everyone is converging, no matter where you start, everyone is converging on the same. We're gonna bring back the human environment because the digital devices have become so divisive, and once faculty started to say we want that, it made it easier to make the social argument about intellectual development. What started as a particularly communitarian goal for our student body very quickly moved into the way faculty think about their classroom.
So on one level, the sort of worry about people disappearing into their devices easily absorbs concerns about AI as well. But there are other you know there are other things going on. Our head of mental health is worrying about students who are over reliant on chatbots for you know relationship advice.
And again, like GPS, if it was just like oh, it dispenses good advice and then people can act on it, that might be one thing. But again, there's good evidence that people who use chatbots for relationship advice are less likely to invest in preserving or repairing a relationship, so it the it literally makes us on average more antisocial when we turn to it for social advice, and that is a deep worry. We have many many fewer ways of intervening directly into students' social lives, of course, and as is appropriate, they are adults. They are coming to us as young adults. These are not children.
Jeff Young:
They're living in New York City, right?
Clay Shirky:
And they're living in New York City, and you know that that feedback loop is a lot harder, right? You can change policy in your classroom, and then you know figure out how to enforce the behavior, and you know faculty Control that the the the people who oversee the dorms can't walk into people's rooms and say, "Don't be asking your chatbot for advice about how to deal with your roommate. Come talk to us. We can we can advise that that's a good strategy, but we do not have the ability to directly intervene in students' lives the way that we do in their educational Experiences, so I wish I had an answer for you, Jeff. This is a hard, long-term problem, but it is also a problem that is going to hit at a minimum every residential college in this country.
Jeff Young:
And I think it's hitting. I think you're. I mean, the MIT report. Back to that. They address this. This is what I focused on in my episode of of the social loss, where students don't do the you know P set groups anymore because they just ask Chat GPT.
Clay Shirky:
And yeah, go ahead. I was just gonna say the terrible thing for us is that we in the classroom when professors in the classroom we make the same. Mistake Silicon Valley makes, which is we tend to over-index on intellect and under-index on emotion, and over-index on individuals and under-index on groups, and so the fact that these tools are having negative emotional effects on groups is something that faculty will deal with the consequences of, but don't immediately see because it's not how we're used to thinking about our students. And that, like, even just alerting people in the institution is a different problem. Like, if I say if students don't do their own work, right, they don't get the learning. Everyone understands that immediately.
Jeff Young:
People, even the students, you say they're bought in. They get that
Clay Shirky:
People were dialed into that problem. I think Ben Thompson's essay on AI homework came out on like the fifth of December of 2022, like less than a week after ChatGPT. I
Jeff Young:
remember that one.
Clay Shirky:
Yeah, essays are appearing saying, "Oh my God, higher education! Like we were at ground zero when that meteor hit. It is harder for our people, particularly our faculty, to see the emotional and social effects.
Even though we all rely on the students having a community and creating norms for each other and teaching each other, we again we like to think we do all the teaching. It's certainly what we're nominally paid for.
But in fact, again, thinking back to your own college career, how much did you learn from your peers? Just, just unbelievable amounts. Often in completely formal situations. If that informal learning leaves college campuses, we are strictly worse off. Even if we change nothing about what the faculty do.
Jeff Young:
Well, I definitely wanted to end it on a little bit of a shift, because you know, as you know, but listeners may not. I got to hang out with you way back in 2010. I came to NYU.
I interviewed you for this big profile that I did for the Chronicle of High Red. It was like a cover story profile for the review. It was really fun. And back then, you had this book that everyone was talking about called "Here Comes Everybody,” and the gist was that the internet and social media, they were going to reshape the world and unseat, you know, who the gatekeepers are, and exciting things were going to happen. And obviously, like 2010, just to remind people, this was a different era. A lot of people, a lot of people were sharing similar hopes and dreams for the technology, and you were looking at things like Wikipedia and saying like we're just going to get more of these like better group projects out of out of this technology. I wonder how you see some of that work now, considering your day job and your the struggles we talked about with AI. Yeah, if you had to write some sort of introduction, how would you?
Would is there a way to square it with today?
Clay Shirky:
I mean, there is a way to square it with today because the you know, I mean, it's funny that book came out in oh eight, so almost almost 20 years.
It'll be you know 20 years in a year and a half. So I overestimated the effect of the internet allowing people to gather in the real world, and underestimated the effect of online mob mentality. Basically, I looked at the index of that book when it was 10 years old, when it was back in back in 2018. I went back and looked the book and clearly look at the index because the index is a kind of an X-ray.
Jeff Young:
Sure,
Clay Shirky:
I wrote more about Meetup than Facebook.
Jeff Young:
Yeah.
Clay Shirky:
I wrote more about the ability of the internet to coordinate this kind of mix of in-person and and and and online action, and less about the social network. The book was written from '05 to '07, just as Facebook was growing. And the thing I missed with Facebook, and I think this is probably the single biggest mistake I made, is not understanding that the social graph. This this now
Jeff Young:
oh that's what they called it. That's right.
Clay Shirky:
Yeah, really large scale. My friends to your friends to their friends to their friends allowed Facebook to get away from the scaling limits of groups because groups, of course, can't have too many members if they stop being groups, right? And so, as long as you're group oriented, you tend to be one of these sort of self-governing human communities. The social graph allowed them to break those limits to scale and to scale to almost unlimited size, but at the effect of eroding social commitment.
And so you start to get what Dana Boyd called context collapse-something you say to your friends suddenly, you know, magnified and broadcast-and Facebook had a huge incentive to overshare because that's where the ad revenue came in.
So the idea that a social media company would systematically undermine human happiness in order to harvest ad revenue was something I missed, and that you know I assumed. People wanted more users because they wanted to serve more ads. That was clear, but it wasn't just that they wanted more users to serve more ads. It was that they wanted those users not to be in coherent groups that knew each other. They wanted them to be in this much larger, much looser social graph formulation.
And one of the earliest backlashes was the rise of the group chat. Right, everybody withdrawing from Facebook and Instagram as their normal social experience, and instead saying, you know, me and anything from a handful to a couple dozen of my friends are hanging out on Discord or whatever.
We got the side Slack channel, just just texting each other.
Even the group chat was a reaction to the social graph, and that you know many of the things I predicted in the book did come to pass. Wikipedia is, I think, still the world's greatest example of distributed peer review. The value of distributed peer review.
Jeff Young:
Even academics have come around to it. Yeah, exactly. Even academics, they used to have. I remember these interesting debates. I think I moderated one between Wikipedia's founder and some academic that was sour about it.
Clay Shirky:
Yeah, no, no, right. All of that. Now, I mean, academics now long for for the days when Wikipedia was the only thing they had to worry about students getting answers from. Right. But the other thing that's the other thing. You know, again, huge success. Linux has become the most important operating system in the world, right? Open
Jeff Young:
source operating are those examples.
Clay Shirky:
You see some of this with the open weight stuff in AI, which is interesting. It's early days, but that's that's interesting work. But the core of the work, and here comes everybody, was really around this idea of the internet letting you find your people and the the imperatives of the advertising industry to get people to pay less attention to each other and more attention to content ultimately won, and I didn't think it was going to, and I was wrong.
Jeff Young:
Yeah, and now we're all living in in that world.
Clay Shirky:
We're all living in yeah, exactly right.
Jeff Young:
Yeah, I do miss the optimism of that time that shaped a lot of my work at the time. I was in total agreement with you. I would say so. No, so I don't. I'm not sitting here blaming you. No, no, no. I look. I just had different. It was an interesting era.
Clay Shirky:
Yeah, I just had this conversation with Paul Ford, another one of those people who like, oh, this is this is an incredible, powerful opportunity, which it was, but it was one in which the you know the the force of extraction overcame the attraction of connection, and here we are. Is
Jeff Young:
Is there anything to learn from that in your work today?
Clay Shirky:
So one of the things I say to about AI all the time is it took 15 years from seeing the harms from social media to the first classroom ban, and it would be good if we didn't wait a decade and a half to do that with AI. We see the harms now. We know what's coming. If we have to wait until the late 2030s to have any sensible regulatory apparatus, we're in trouble. So that so we might see
Jeff Young:
more we might see more phone free classrooms and device free classrooms because that takes AI out,
Clay Shirky:
And funnily enough, the phone free classrooms and device device free classrooms are downstream of the social media and and and smartphone stuff, and it's easier for AI to ride along with that, whereas the social media stuff had nothing to ride along with. There was no familiarity with what does it mean for a whole class of software to be potentially bad for teenagers. Like that, just you know, video games is the thing people most worried about, and most of those worries were misplaced.
And so, for a long time, the lesson was, "Oh, the video game thing, the rap music lyric thing, whatever. That was that was a that was, you know, whatever. It just wasn't worth getting that worked up about. So it took people a while to realize, oh, social media actually is worth getting worked up about. Right? Maybe I think optimistic. The optimistic story is now being alert to those worries. We can move faster on AI.
Jeff Young:
Well, I really appreciate it. I hope to check back in with you and you know in six months. See you,
Clay Shirky:
Dr. Cross too. Yeah, yeah, it was great.
Jeff Young:
Oh, thank you so much for doing this. Good to talk to you.
Clay Shirky:
All right, good. Take care.
Jeff Young:
This has been Learning Curve.
I'm exploring how education is adapting now that AI seems to be everywhere. I'm eager to hear what you think of Clay Shirky's take. Maybe you're a student or professor trying to figure out all this, and you have a story to share.
Reach out to me at jeff at learningcurve.fm.
You can find more about my journalism and public speaking at LearningCurve.fm. I hope you'll follow Learning Curve podcast so you don't miss any future episodes, and tell a friend about the show so we can continue to grow.
Our theme music is by Komiku, and the episode art was generated by Midjourney.
Thanks again to this episode's sponsor, Studiosity.
Until next time, thanks for listening.