What does a postsecondary educational institution that is of genuine value to students in a world with omnipresent AI look like?
One vision is essentially a full embrace of the technology as a layer on top of the existing structures, under the assumption that AI will be a totalizing force. This appears to be the philosophy of places like Ohio State, which has declared an intention to infuse AI competencies into every course. I have registered my skepticism regarding this approach previously and honestly don’t have much to add that I haven’t said before.
That totalizing vision is predicated on what I believe to be the dominant present view of what college is for—namely, return on investment as measured by your immediate prospects for employment requiring a degree immediately out of college. I think this is a lousy measurement of even that narrow definition of ROI—our first job is not determinative of our lives, thank goodness—but it remains pervasive.
But those of us who have graduated and gone on to other things recognize that the returns of college go well beyond our potential to land that first job.
An alternate vision for higher ed comes from filmmaker Ken Burns, who on a recent appearance on The Chronicle’s College Matters podcast talked about the importance of his years at Hampshire College (which will close at the end of this year) to his life and career.
Burns says to host Jack Stripling, “The greatest danger that faces higher education is this transactional tendency that everything is about an exchange that ultimately is economic and not what education should be, which is transformational,” to which I say, “Sing it, my man.”
He says that his time at Hampshire “rearranged all my molecules.”
Rearranging molecules actually sounds like a pretty solid by-product of time spent at a higher education institution. Stripling’s interview with Burns suggests that the uniqueness of Hampshire had something to do with this rearranging of Burns’s molecules, but as the graduate of a bog-standard state flagship (University of Illinois at Urbana-Champaign), I can testify that a goodly number of my molecules were indeed rearranged by my four years, even though I was an often uninspired student sleepwalking through a nontrivial number of my classes.
While Hampshire was in the explicit molecule-rearrangement business and UIUC was not, the mere fact of being in community among others whose molecules were also being rearranged, and having the good fortune to happen across a handful of instructors, combined with a number of experiences (such as meeting the woman who would become my wife) that challenged my previously held notions, my stuff was shifted around a pretty good amount, too.
This may seem strange, but the University of Illinois’s apparent indifference towards individual students was a great aid in this rearranging. To make something interesting out of my life, I had to figure out what was meaningful … to me. I didn’t even have an academic adviser after my sophomore year, let alone some kind of close guidance about how to map a path toward success. I had to decide what success was.
I also had the good fortune of graduating college toward the tail end of the world not caring overmuch about one’s GPA. My 3.1—shameful by today’s standards, my students have assured me over the years—never proved a hindrance going forward. I got that first job (paralegal), went to graduate school (two more “worthless” degrees, M.A. in literature, M.F.A. in creative writing), got a real-world job (market research consultant), got some traction writing and so on, and so on.
Our lives are really a series of “and so ons,” where exogenous economic events—like the arrival of generative AI—may force us to rearrange our molecules to fit the moment. I’ve found my liberal arts degrees well suited to the challenges of the last 30-plus years. ChatGPT has been both a scourge and the best thing to ever happen to my career.
Another possible proposed future for higher education comes in the form of an interview of Hollis Robbins at The New Yorker by Jay Caspian Kang, which theorizes a future of higher ed akin to the network of OnlyFans, instead it would be “OnlyProfessors,” where individual students seek out very specific expertise online after having mastered a more general curriculum via interactions with AI. The professor would be the source of the credential, not the institution.
Robbins is a polymathic, idiosyncratic, sometimes brilliant thinker, but this is truly crackpot stuff that even Kang, someone who enjoys cuffing around the elites for their shortsightedness, cannot help but express some skepticism toward. Unfortunately, it is the kind of crackpot stuff that gets the moneyed interests of Silicon Valley and a handful of academics convinced of their own unique brilliance salivating.
I suppose there is some potential for molecule rearranging in this model for the minority of students who arrive at age 18 being able to learn through interactions with AI models, but I believe the opportunities for the collision of atoms leading to interesting recombination of molecules is greatly enhanced by a richer and more diverse arena of possibilities.
Robbins has long been bothered by the “generic” nature of higher education intuitions, a condition fostered through regulations and requirements around courses and learning outcomes, and I take this point, but there is nothing inherent to the larger university that requires the student experience to be generic. To the extent that the experiences of students are reduced or constrained, it is a by-product of believing that an education must be optimized, rather than experienced.
I went to a generic university 35 years ago, and my experience was anything but, primarily because it was mine. In my years of working with students, one of the chief challenges I confronted as their writing instructor was a belief that the world was not interested in what they had to offer beyond their capacity and willingness to secure the college credential.
Robbins hypothesizes that 60 to 70 percent of faculty will lose their jobs to generative AI, and the only faculty that will (or should) survive are “specialists” who will have the capacity to do things AI can’t, such as be an expert in the African American sonnet tradition—Robbins’s specific academic expertise.
The test for the faculty who get to stay will be “What specific knowledge do I possess that AGI does not?” Only those “with concrete defensible answers” will have a place in this new postsecondary education world.
I had to think about this for a moment and realized that under Robbins’s criteria, if stacked against hypothetical AGI, I cannot imagine possessing academic knowledge beyond its capacities.
What I “know” that AI does not are things like how to probe a student’s response when they seem confused, or when and how to offer a possible diagnosis for something they are struggling with in their writing. I know how to structure assignments and assessments in a way that stands at least some chance of fostering engagement.
By the end of a semester, I would know all kinds of things about my students than an AGI would not, because I have been in community with them for 15 weeks, but this is the kind of thing that Robbins is generally dismissive of, saying that “experiences” are “hollow.”
I would know what they find funny through the rhetorical analysis on a work of humor they would write for me. I would know what makes them angry based on the researched essay they would write from my question “Why am I so angry and what can I do about it?"
I would know something about their taste in music based on what they would recommend I play on the computer/projector prior to class. I would likely know where they’re from, what they’re majoring in and whether or not they’re genuinely interested in their major (often not).
I would know who arrives sleepy, who arrives hungry, who arrives sneezy and grumpy.
I suppose an AGI with access to the data in a learning management system (provided it hadn’t been hacked and held for ransom) would possess some of these facts, but it would not care about any of it.
A single first-year writing course cannot compete with the four years of a Hampshire College education that Ken Burns experienced, but I think the mere fact of a group of humans gathering together to work on similar challenges alongside each other—as happened in my humble first-year writing course—may have shifted an atom or two or 12.
Having now spent over two years writing about teaching and AI and traveling to campuses to speak on the subject, I’m more convinced than ever that universities do not need wholesale reinventing for the AI age, so much as a collective, thorough reconsidering of what kind of learning gives students what they need to move forward, step by step, through our ever-changing world.
