The other day, I went into a bookstore. There was a section on “classics,” which was essentially a section of works of nineteenth and twentieth century literature.
I saw volume after volume of books I’d read before in my youth, from The Brothers Karamazov to The Plague, and as I did so, I thought to myself, “Man, I could never read these books now!”
Part of the problem was physical. When I was young, I always smelled books. I would open a book, hold it up to my nose, and inhale deeply. I wanted the book to feel good in my grip and to have that smell, whatever it was – a certain kind of ink on good paper?
Oh, and the smell of Japanese literature books!!!
However, in looking at the books that I saw in this bookstore, I didn’t even try to smell them. I could see right away that the paper was poor quality, the binding was weak, and the print was too small for my failing eyesight.
So, physically I did not want to have contact with the books, but also at an intellectual level, I just couldn’t imagine myself being able to conjure up the mental strength and patience to get through a novel.
There is a history professor at Princeton by the name of D. Graham Burnett who published an article in The New Yorker a few months ago entitled “Will the Humanities Survive Artificial Intelligence?” with the subtitle “Maybe not as we’ve known them. But, in the ruins of the old curriculum, something vital is stirring.”
It’s an interesting article that imagines a future for the Humanities where people wrestle with irreducibly human questions, such as how to live and what to value, questions that some would argue machines can’t answer (although many people are asking LLMs precisely about these issues. . .).
I found this piece to be quite elitist, as I couldn’t imagine such a future for the Humanities outside of a very small circle of (wealthy) students at elite universities like Princeton.
More recently, however, Burnett was interviewed for the podcast, Hard Fork, and in this interview, he gives more of a sense of what he thinks the world beyond that tiny circle of elite students will look like.
However, even for the elite, he acknowledges that “long-form immersive literacy,” namely, the reading of books like those that I saw in the bookstore recently, is a practice that “is coming to an end” (personally, I think it has been dead for some time now, but. . .).
I thought I would share this interview here. I think it’s worth “listening” to, and I enjoyed it more than “reading” his article.
I’ve already accepted that most university students in the humanities are incapable of reading anything more than a few pages long. Last term I gave a three page reading one week and nobody even bothered to look at it. As the inability to read or reason becomes ever more widespread throughout my own society, I have shifted from thinking about ways I could try to prevent the slide into illiteracy and ignorance to thinking instead about how I could remove myself from the world but still survive and continue to enjoy reading my books while everyone else continues on down whatever path they choose. I still haven’t managed to come up with a good way of doing it yet, so suggestions would be greatly appreciated.
Perhaps because I have spent the past several years teaching Southeast Asian history at an allegedly “elite” US institution, my experience with undergraduate readers has been much more positive. I have assigned “This Earth of Mankind,” “The Coroner’s Lunch,” “Letters from Thailand,” “Lord Jim,” Zhou Daguan’s “Zhenlafengtu ji, “”The Year of Living Dangerously” and other long works of in their entirety. And other works like the “Travels of Ibn Battuta,” or the writings of Isabella Bird and Anna Leonowens in shorter measures. We have spent two weeks or four class sessions on the longer works, discussing them in detail, with the students responsible for presenting and analysing the texts (after the provision of some scaffolding). These have mainly been in “first-year writing seminars”, capped at 15 students, and so the teaching has been boutique. But it has also meant that in senior seminars I have been able to assign all of Dian Murray on the pirates of the South China Sea, or all of Carl Trocki’s “Prince of Pirates” of Warren’s “Ah-Ku and Karayuki San”. I should say that I did the same thing when I taught at a much less self-avowedly “elite” state school in the south. Students in both places were serious, engaged, and showed evidence of deep and thorough familiarity with the texts (on which they were tested). I treated my classes like classes in the engineering faculty, in so far as it was possible to fail, a deployed the full range of grade, the homework was very demanding in terms of time and engagement, and I expected a lot of “extra research” (ie Googling and Wikipedia-ing) about things in the texts that might be unfamiliar (what is a prahu? where is the Songkhla peninsula?), I would cold call students, and it was perfectly possible to fail assignments and the class as a whole. I never had to cancel a class for want of enrolment. Understanding Southeast Asia is every bit as serious, difficult, and intellectually demanding as cellular and molecular biology or quantum electrodynamics. I would even argue more so. There were no elementary-school level “map quizzes” or “ID questions” since I expected students to learn these things, the way electrical engineering students are expected to learn things like the Heaviside functions, en passant, to solve other more challenging problems. So “getting an A” in “Southeast Asia – Decolonization and Independence” on campus meant as much as getting an A in “Quantum Physics III”. My experience is that if students think that they will actually gain valuable perspectives, skills, understandings, and “bildung,” if the subject is treated seriously and students are held to account, then students will take the work seriously and do justice to difficult and complex texts.
What I find troubling is, if it is true that -in general- students will no longer engage with long form texts,(and my own experience is set aside as an outlier, a fact I am happy to accept), is just how much deep and genuine pleasure, quite unlike any other, they are missing out on. Booze, and weed, and video games and sex are great – but so are hours spent in different imaginative worlds with people long dead or whom one will never met. It is sad and strange to me that future generations might never fully know this pleasure at all.
In the early 2000s, I would have classes with 5 papers (basically one every 3 weeks), each of which was based on a book-length reading, like This Earth of Mankind or a memoir like First They Killed My Father (which I used to have them also read a critique of and then write about both). And yes, it was clear that many got pleasure from “hours spent in different imaginative worlds with people long dead or whom one will never met.”
By the early 2010s, however, a lot of students just couldn’t handle that anymore. Part of it was that they were not as willing to read, but perhaps a bigger part of it was also just that they had to work 7 jobs and didn’t have time. So I stopped teaching that kind of course and focused on working with primary sources in digital archives. That worked pretty well, and in some incarnations the course was quite demanding in that it required constant effort. However, after COVID, the ability of students to do that collapsed.
Geography quizzes – I’ve always done them. Again, in the early 2000s, most students would get 17-20 out of 20. By the 2010s, the norm was like 8-12 out of 20. By the early 2020s, I would give the quiz, most people would get 4-6 out of 20, and I’d say, “Ok guys. Let’s try this again! I’m going to give you THE EXACT SAME quiz in the next session, so get ready!!”
The result? One or two would improve, but the rest would still get 4-6 out of 20. . .
All of that is before AI really started to take hold.
This book, “AI Snake Oil,” by two Princeton computer scientists is particularly good:
https://press.princeton.edu/books/hardcover/9780691249131/ai-snake-oil
And it is concisely discussed here:
https://press.princeton.edu/ideas/the-truth-about-bullshit-on-bullshit-in-ai
Thanks for the comments/links!!!
So, I can’t listen to this podcast because it is too slow. My brain needs words to pass through it at a faster rate for me to have the patience to listen. I listened to 51 seconds of this podcast, and finding no way to increase it to 2x speed, I gave up.
That may seem crazy to you, but I think my reaction is that of what is now a “normal” person and is also an example of how the academic world is out of touch with reality (my kids and their university classmates watch their profs’ recorded lectures [which the university demands that they record and upload] at 2x speed).
I asked Grok to provide me a detailed summary of the book, which it did instantly.
From that summary, I think this “AI is snake oil” argument is itself a form of “snake oil.” There are all of these unrealistic hyped expectations, and then people pointing out that AI doesn’t achieve those (unrealistic) expectations, but it doesn’t have to do that to have enormous negative consequences on certain realms/professions.
LLMs can produce texts that are better than the ones that most of our students produce. Our “purpose” is to teach students how to do that. . . That’s all it needs to be able to do to wipe out our profession, and it can already do that. It doesn’t have to do any of the things that these authors talk about.
But over time, it will get better at doing some of those things. And when you train more specific models, rather than using the general LLMs, it will also get better at whatever task you put it to.
Ok, so this morning I found the podcast on YouTube and was able to speed it up. What I heard confirms what I thought. The main critique here is about “predictive AI.” The guest doesn’t say much about “generative AI,” which is the form of AI that affects our profession. And I would make the same point: Predictive AI doesn’t have to be good for our profession to be screwed. As long as generative AI is good enough to write a paper, we’re screwed, and generative AI is good enough to do that, and has been able to do that for a while already.
I pulled out a couple of key quotes here:
1) Guest: “What a lot of the book is about, though, is predictive AI, a very different kind of AI. This is not AI that we interact with on a day-to-day basis. This is AI that’s used by companies and governments often to make decisions about us in the background. So when we apply for a job, our resume might be uh analyzed by you know an AI system. In in healthcare, these algorithms are often predicting how long a patient needs to be in the hospital or you know various other things related to healthcare.”
2) Host: “If there’s one parting thought that you could leave listeners with in just terms of how they think about AI, how they go about and interact with it, um, is there something that you’d want people to take with them and remember when they’re using the chat bots or when they’re using any AI product?”
Guest: “So when we’re talking about generative AI, right, as opposed to the other more hidden types of AI, whether it’s predictive AI or things that industries use to optimize, you know, their warehouses and backend operations, those are types of AI that have been used for decades. Generative AI is new. It’s an area where there’s a lot of hype. But there’s one very, very good thing about it, which is that listeners don’t need to listen to, you know, any quote unquote experts or pundits out there, including me. They can come to conclusions on their own by using uh these tools.”
– Coming to my own conclusions by using these tools is precisely what I’ve done, and is why I am so adamant that our profession is in serious trouble.
Re: D Insor (I promise that I’m not internet stalking you, just catching up on some of LMK’s posts)
“capped at 15 students, and so the teaching has been boutique.”
Roughly what percentage of all undergrads in the USA are enrolled at elite schools that regularly offer this kind of experience? At my tuition-dependent employer, which is closer to the norm for four-year institutions, enrollment in most humanities/social science sections is capped at 35. Students only register for these courses only because they fulfill general education requirements in the traditional 8-semester, 120-credit model of the bachelor’s degree. Now we are starting to see a shift toward 90-credit B.A. for programs oriented toward specific occupations — no humanities/social science gen ed requirements at all.
The commentators who work at the Harvard-Yale-Princetons of the world saying “everything is fine” don’t
pay attention to statistics.
1) The students in the courses I taught in a state school were usually about 80% non-History majors. They were students in other departments who took the course as an elective to fill a Humanities requirement.
When I saw the number of majors decline rapidly in the 2010s, I told my colleagues, we are going to be a department that serves people in other departments, and that’s what I think it and probably many other history departments have indeed largely become.
2) Looking at enrollment patterns, courses that are offered fully online fill up the fastest and can get the highest enrollments if they are not capped.
What does that tell us? Many students who need a Humanities course to fulfill an elective requirement want to take whatever is perceived as the most convenient and hopefully easy (i.e., an online course) to do so (I have yet to see a comment anywhere by a student praising an online course for its rigor).
So, the comment you made to another post: “Therefore I disagree with people like Ted Gioia, who is a smart guy, but who thinks that the way to counter AI is to make universities more like Oxford in the 13th century (https://chadraymond.substack.com/p/the-death-of-curiosity).” – Yes, I completely agree.
As such, we have a system where fewer and fewer people want to major in certain fields, and a significant % of the people taking courses in those fields are people who just need an elective fulfilled, and a certain % of those would prefer to achieve that with as little hassle/effort as possible.
AI tools are obviously going to find a welcome reception in such a context, but the point you make here is also really important: What happens when such a university decides to reduce the credit hours from 120 to 90 and that Humanities elective is reduced or eliminated?
“Oh, and the smell of Japanese literature books!!!”
Oh no, professor, I’m having a Mishima flash-back, and it’s all your fault, hahaha…
When I was 12 years old (not long after the ‘War & Peace’ debacle), I was wandering around the U of M campus and/or the State & Liberty area of downtown Ann Arbor one day, as I often did, and I happened to read something about Yukio Mishima (I think it was the Ann Arbor Film Festival schedule, which was more like a little free magazine). I’d never heard of him before that day, and he seemed really interesting- as I remember, he was described as both “notorious” & “eccentric”, which immediately piqued my interest. I raced over to The Dawn Treader Book Shop to ask the proprietor, Bill Gilmore, if he had anything around that might enlighten me further. Bill started cracking up, and said he would try to help, but that I shouldn’t tell my father he had done so as he would most likely be pretty annoyed with both of us…
I followed Bill around the subterranean maze stuffed with books and papers that functioned (after a fashion) as his store, and he gave me a fairly extensive précis of Mishima’s life and career, which ended something like this: “After all the time they spent poncing about in their pink uniforms, the Shield Society just weren’t very competent samurai. In the end, they botched up his otherwise perfect act of seppuku by not being able to properly cut off his head!”…
Needless to say, this only increased my enthusiasm & curiosity!
Bill couldn’t find anything relevant that day, but a week or so later I was at the Ann Arbor Public Library, and discovered that they actually had a copy of one of his most famous books, ‘The Sailor Who Fell from Grace with the Sea’! I was so excited that I would be able to read his work! I found the already aging hardcover on the shelf, took it down, and opened it up. As I did so, I noticed an odd and unexpected smell, so I warily raised the book toward my face and took an exploratory sniff, which I immediately regretted.
I nearly started coughing and choking due to the smell being so strong. The book didn’t have any obvious stains on it, but it smelled as though someone had taken a jar of spoiled mustard, dumped it on the floor of a sodden, filthy pissoir, then used the book to carefully stir the resulting mess for some time. Like I said, the book looked just as clean as any other book of similar vintage which one might find on the shelves of a busy public library, but it had a stench that was just unspeakably vile. So I checked it out and took it home so I could read it. Of course I did! I also, however, took the precaution of disinfecting my hands every time I touched it.
This probably won’t surprise anyone who has read it, but I was very disappointed with that book by the time I finished reading it. It was not by any means the worst book I ever read, but…
More than thirty years later I read ‘Patriotism’, which I absolutely loved (and the film is as good as the book!). The two forms of ‘Patriotism’, taken together, may well be Yukio Mishima’s masterpiece; but I’m still bemused by the fact that so many people seem to like ‘The Sailor Who Fell from Grace with the Sea’.
Even all these years later, every time I think of that book I sense a faint whiff of spoiled mustard and old piss…
Ha! Great story!!
Yea, I’ve always wondered if what I enjoyed smelling was just something toxic in the paper or ink. . .
I grew up up-wind of a paper factory. So maybe over the years whatever that southern breeze delivered to my nose had some kind of deranged effect on my brain. . .
I’m so glad you enjoyed it, professor!
We had a paper mill in one of the places I lived as a child, too- you may have a point! I still remember how horrible it smelled while driving past on the way to my great-grandmother’s house.
For the record, though, I also have usually enjoyed the smell of books. If you spent as much time in libraries & bookstores while growing up as I did, I suspect those associations alone could explain it for both of us.