Many within higher education have been watching the advancement of artificial intelligence (AI) models with amazement. And fear.
Some fears are existential: AI will kill us all. Some are academic: AI will facilitate student cheating. And some are financial: AI will displace well-paid and, until now, secure tenured jobs for faculty by rendering faculty and universities at large obsolete. Examples of writers grappling with this possibility include
- Hollis Robbins: “In 10 years, when students can get personalized content mastery for free or cheap and arrive at college having completed what we currently teach in years 1-2, what exactly are we selling? Research apprenticeship? Network access? Credentials? Time to mature? What’s the honest answer?”
- Rikki Schlott: “We’ve entered the age of AI — yet parents are still ponying up small fortunes for their kids to learn jobs that will soon be antiquated (if they’re not already).”
- Ted Underwood: “higher education risks losing its central place in the production and distribution of knowledge. It isn’t hard to envision a world where everyone knows that you get intellectual substance from an AI subscription, while formal degrees confer prestige mainly through association with beautiful campuses and victorious football teams.”
I can’t say much about whether AI will end civilization or solve the perennial problem of student cheating, but I do have an unorthodox view on the question everyone in academia is whispering about: will AI make faculty—and the university itself—obsolete?
[RELATED: Universities Are Racing Toward AI. Is Anyone Watching the Road?]
The short answer is yes. AI will make universities obsolete. But that’s hardly fatal. After all, higher education has outlived its own obsolescence before, and it will almost certainly muddle its way through this one, too.
The long answer is the following:
Universities and their faculty have been functionally obsolete for centuries. The traditional university lecture was created because books were prohibitively expensive in the past. A thousand years ago, books were written by hand and were therefore exorbitantly expensive. Economist Brad DeLong estimated that a book would typically cost about $50,000 in today’s dollars, which in turn meant that a college-level education would cost about $1.6 million just for books. The lecture was the solution to this problem. As DeLong wrote,
you assemble the hundred or so people who want to read Boethius’s The Consolation of Philosophy in a room, and have the professor read to them–hence lecture, lecturer, from the Latin lector, reader–while they frantically take notes because they are likely to never see a copy of that book again once they are out in the world administering justice in Wuerzburg or wherever…
But the invention of the printing press in 1440 changed all that.
While it took some time to develop fully, the printing press made books cheap. For example, today, I can buy Boethius’s The Consolation of Philosophy on Amazon for $9.91. In this case, the printing press reduced the price of this book by 99.98 percent. In other words, the lecture and the universities that sprouted up to provide them have been obsolete for centuries. And yet lectures and universities persist to this day.
[RELATED: AI’s Higher Ed Takeover Is Not Inevitable]
There is an interesting question of why universities have survived the emergence of the printing press.
DeLong offers some theories, as does Bryan Caplan. But for our purposes, what matters is that universities did survive. And for precisely that reason, they are likely to survive AI as well. Indeed, with only a few tweaks, the quotations at the beginning of this piece could just as easily be describing the arrival of cheap books rather than AI—a reminder that easier access to knowledge has never been a fatal blow to universities.
In fact, the university is in an even stronger position today than it was when the printing press was invented because demand for its services is artificially inflated by both credential inflation, which tends to require postsecondary education for jobs that don’t truly need it, and massive government subsidies that make higher education seem more affordable than its true cost.
The bottom line is that if universities survived when a college education could be had for the cost of a library card, they’ll survive when one can be had for the price of an AI subscription, too.
Follow Andrew Gillen on X.
Image by Sweeann on Adobe; Asset ID#: 917190692
““In 10 years, when students can get personalized content mastery for free or cheap and arrive at college having completed what we currently teach in years 1-2, what exactly are we selling? Research apprenticeship? Network access? Credentials? Time to mature?”
Sixty years ago, the average “college prep” high school graduate already knew most of what colleges now teach in “years 1-2” — and that is without mentioning the vast amount of remedial instruction being done in higher education today.
What we essentially have done is gone from an 8th Grade education a century ago to a 12th Grade education sixty years ago, to a 16th Grade education today — to be a “respectable” member of society, to have and hold a “good” job today, one must have a college degree.
I pick sixty years ago because of the 1965 Higher Ed Act — sixty years ago is when the Federal money started flowing into higher ed. And a century ago — the 1920s — is when we started to see an expectation of taxpayer-funded high schools although the expectation that this be universal really didn’t arrive until after WWII.
Remember that parents always could pay tuition for a child to attend a private high school (often called an “Academy”) much as parents could also pay tuition for a child to attend college. The changes were that the taxpayer paid for it, and the expectation that all should be permitted to attend.
It’s not gradeflation as much as a lowering of admissions standards necessitated by the presumption that everyone ought to go to (and graduate from) first High School and then College as well. This corresponds with a related decline in what students knew first upon graduating from the 8th grade, then from high school, and now from college.
Conversely, membership in the middle class has gone from requiring an 8th grade education to a 12th grade education, to now requiring a 16th grade education. In terms of knowledge, we’ve just dropped the quality of the 16 (actually 18) year education down to what the 8 year education used to be.
If AI (or Martians, or whatever) were to arrive and start conveying sufficient information to raise cohort knowledge levels back to what an 8th grader knew a century ago, much of higher education as it exists today will be kaput. And as to the third rail — research — the very real question is if it is of sufficient value for society to want to continue to fund it.
Let me rephrase that — regardless of the economic system, capitalistic, communist, or whatever — allocation of resources is related to the perceived value of the enterprise. Likewise, the willingness of people to participate in the enterprise is a combination of financial rewards and other rewards — for example, a lot of rural ministers drive school buses or have a heating oil delivery business because their small church can’t afford to pay them a living salary.
Right now, undergraduate tuition subsidizes research — both directly and indirectly (Federal funding to research could instead go to Pell grants). Maybe the taxpayers would be willing to continue paying for research into the sex lives of lesbian African tree frogs, but I don’t think so…
And maybe those researching said lesbian African tree frogs are so dedicated to their research that they are willing to drive home heating oil trucks on weekends to pay the bills. But both having driven a home heating oil truck and knowing researchers, I also don’t think so…
So what will higher education HAVE to offer???
And the other thing to bear in mind is the social disruption caused by the 1964 Civil Rights Act — no longer were the good jobs restricted to White males, and bad things happened to companies that attempted to continue their existing discriminatory practices.
Imagine a similar law based on social class. Imagine an environment where a different form of affirmative action excluded the Harvard graduate, where the network that mattered was amongst CDL holders who had driven said home heating oil trucks.
When you look at the visceral hatred that half the electorate has for Harvard (et al), well, stranger things have happened — eg the 1964 Civil Rights Act….
” demand for its services is artificially inflated by both credential inflation, which tends to require postsecondary education for jobs that don’t truly need it, and massive government subsidies that make higher education seem more affordable than its true cost.”
Currently artificially inflated. And this will not continue.
Credential inflation is currently caused by the asinine Griggs v. Duke Power Co. (401 U.S. 424) decision and the even more asinine EEOC interpretations of the “disparate impact” concept of discrimination created by this decision.
Like Roe, Griggs will eventually be overturned and this, too, shall pass.
And as to the government largess, can you not see the writing on the wall?
What AI will do is eliminate the value of alumni networks in job placement and that will be the utter end of higher education as we know it.