I Was a University AI Czar. I'm Not Equipped to Teach in the Age of AI.

Over the past academic year, I served as the Inaugural Faculty Fellow for Artificial Intelligence (AI) at my university, a mid-sized public institution in the South. In that time, I acted as an AI curriculum developer, community resource, media personality, and intermediary between faculty and administration. I orchestrated a free AI certificate program that boasts nearly 50,000 enrollees from 150 countries; co-designed an interdisciplinary AI minor; co-led a task force that developed a workflow for the procurement of new AI tools; gave 30+ talks across 4 countries; wrote a case study about the strategic deployment of AI at my university; and advised non-profit organizations, universities, and elected officials on AI strategy and policy. It has been a wonderful journey that resulted in a promotion, countless new connections in the local tech scene, and solidifying my university’s reputation for preparing our students to join an increasingly AI-powered workforce.

Despite these accomplishments, there is one inescapable conclusion I reached from my time in the trenches—I’m not equipped to teach in the age of AI.

How did I arrive at this self-deprecating destination? Allow me to explain.

I would consider myself a good teacher. I have nearly 20 years of experience as a classroom instructor at the college level (as a graduate teaching assistant, adjunct, tenure-track professor, and tenured faculty member). My overall instructor effectiveness rating as a professor is solid (4.6/5.0 over 12 years), and it has only increased throughout the stages of my career. I have won multiple teaching awards. I am a certified online course designer and have had numerous courses pass successfully through internal Quality Matters (QM) review. In short, I’d like to think I know what I am doing when it comes to providing instruction.

And yet, AI has laid bare just how woefully ill-equipped I am to meet this moment in higher education. To make one thing clear, it’s not because I am fundamentally incapable of adjusting my teaching in response to this technological disruption, but rather because I just don’t have the time and energy necessary to rise to the occasion.

---

If you’ve been following the discourse in higher education about teaching and AI, you’ll know that the best instructors around have largely settled into two diametrically opposed camps. The first camp consists of AI Enthusiasts. These folks are innovation-minded experimentalists who saw AI for the tidal wave it was and decided to surf it. Accepting that faculty aren’t going to effectively halt student AI use on or off campus (indeed, 80-92% of students report using AI), they offer an array of frameworks, techniques, and strategies intended to help instructors employ AI productively and responsibly in the classroom.

The second camp consists of AI Resisters. People who fall into this group believe that AI is eroding the human capacity for critical thinking and writing, so the only option is to fall back on tried-and-true methods of assessment, including in-class hand-written essays, oral exams, and ploughing fields with brute force (ok, so maybe the last one is a slight exaggeration). They devote substantial time in the classroom to cultivating humanistic skills and a culture of AI abnegation (often at the expense of covering disciplinary content). The thinking here is that these are the virtues worth developing in a world saturated by a soul-crushing technology that will steal all of our jobs, destroy the environment, and deprive us of the beauty, goodness, and truth offered exclusively by an analog life.

To be sure, there is a large contingent not often mentioned, an Exhausted Majority (adapted from the Hidden Tribes report), which has some minimal exposure to AI but isn’t quite sure what to make of it just yet. You rarely hear from these fence sitters; they are waiting to see how it all shakes out before taking action.

To which group do you belong? [Image generated by ChatGPT]

The reason that I claim I am not well-suited to thrive as an instructor in the age of AI is because both AI Enthusiasts and AI Resisters put a lot of thought and energy into completely redesigning their classes in response to AI. This is the one takeaway that I don’t think the Exhausted Majority has fully accepted yet—to excel as a teacher in this AI era, you need to totally revise how you teach and how you assess what students learn in your classes.

---

Since ChatGPT arrived onto the scene in November 2022, I have taught very different types of courses—fully asynchronous online, fully in-person, and hybrid. My experiences teaching each of these classes expose the gulf between traditional pedagogy and a pedagogy fit for AI.

As you might imagine, my approach to AI might be characterized as “permissive.” That is, recognizing that students will likely make use of AI in the absence of my instituting a draconian surveillance regime, I do not ban it in any of my classes. Turning the classroom into a digital panopticon cuts against my values as a teacher. I’d rather my students cheat than I devolve into a private investigator charged with sniffing out academic criminals. Instead, I allow students to utilize AI within predefined parameters, which include explicit disclosure. I include a fairly robust AI policy in my course syllabi.

In these three courses, I have witnessed remarkably different student behavior. In the fully synchronous online course (which passed QM review and took me an entire semester to create), unattributed AI use has almost certainly risen over time. I admittedly don’t have an answer for this; it seems predestined given the modality and types of assessments I use (i.e., short discussion posts). My fully in-person course was an absolute joy to teach, and it brought me back to my origins as an instructor, relying on student presentations, thoughtful provocations, and the kind of real-time classroom engagement that is the stuff of Hollywood films ("O Captain! My Captain!" anyone?).

But my hybrid course was a revelation in its own right. Most of the content (including short module podcasts narrated in a digital clone of my own voice) was generated using AI, but our activities inside and outside the class enlivened otherwise banal subject matter (i.e., writing resumes, networking, etc.). Students engaged in AI-generated simulations and role-playing exercises, experts provided lectures and shared their experiences on career panels, and several classes were spent entirely outdoors as part of short field school excursions. One student in that class told me the following semester, “Your class made me look forward to Mondays.” High praise for a professional development course!

These experiences indicated to me that while there remain productive ways to integrate AI into the classroom, online courses seem especially plagued by the technology. To this day, I do not know how universities will be able to offer entirely online degrees without either 1) imposing severely draconian surveillance measures to keep pace with the need for quality assurance, or 2) tacitly accepting the erosion of academic integrity. Meanwhile, the backbones of our online courses, learning management systems (LMSs), have been exposed for their vulnerability and precarity, as evidenced by the recent successful hacking and extortion of Canvas.

I can say this much—whatever solution our industry comes up with, it’s likely to emerge from teaching and learning centers. Contrary to what Paul Schofield wrote in the Chronicle of Higher Education, pedagogy experts are the best hope we have to equip today’s faculty with the tools required to succeed in this uncertain educational environment. As I always tell my students, “I was trained for 7 years to become a researcher and 2 days to become a teacher.” The idea that only disciplinary experts know how to teach and have nothing to learn from so-called “nonscholars” is so laughable that one has to wonder whether an AI agent jokingly wrote that sad opinion piece to troll the whole academe.

I have also realized that, for higher education to succeed in this brave new world, we will need thoughtful integration of the pedagogical styles of both AI Enthusiasts and AI Resisters. In other words, I think colleges and universities are going to need some spaces that are AI proof (to the extent that is possible at a time when wearables like AI glasses threaten to justify suspicions of the near- and far-sighted alike) just as much as they need contexts where tinkering with AI is welcomed.

But above all, both of these approaches will require wholesale rethinking about how we provide instruction and how we determine whether or not our students have met our lofty learning objectives. The central problem, as educators like Jason Gulya have noted, is the fundamental misalignment between what faculty want students to gain from their classes and what the market signals to students about the import of a college degree (for another excellent critique of the transactional model of education, check out this post by George Cusack and Jennifer Ross Wolff). I plan on writing about this aspect of meritocracy in a future post.

---

It is no longer sensible to operate under the assumption that we are training and graduating students who will go forth into the world in need of exactly the same skills we obtained when we were in school. I dispensed long ago with the idea that my students need to learn how to produce a 20-page research paper on an arcane topic in order to credibly claim they had received training in the field of political science. As much as I respect and appreciate my colleagues in the humanities and social sciences, I think we would be wise to drop this conceit. That does not mean, however, that we need to abandon the wisdom of our disciplines in service of a universal, totalizing acceptance of technology. One of my favorite thinkers about AI in education, Nick Potkalitsky of the Educating AI Substack, has developed an approach he calls DSAIL (Disciplinary-Specific AI Learning) that does an excellent job demonstrating how we can harness our unique disciplinary expertise while also preparing our students to be (responsibly) AI literate. We can walk and chew gum at the same time.

I agree with those who have argued that we cannot anticipate what skills will be necessary or what jobs will be like in the future. But I also agree with others who have contended that if college is to mean anything, it must deliver a kind of experience one cannot receive outside of it. That experience should include deep thinking, a willingness to learn about and debate life’s big questions without fear of reprisal, and practice articulating one’s thoughts through the written and spoken word. It should also include access to and the ability to experiment with emerging technologies, coupled with humanistic reflection about those same tools. I fundamentally believe in the inevitability of AI, and I feel it is the duty of higher education to navigate its arrival, not deny its existence. We will not cast aside our AI tools any more than we have thrown away our calculators or prevented students from contacting professors via email.

“I fundamentally believe in the inevitability of AI, and I feel it is the duty of higher education to navigate its arrival, not deny its existence.”

Higher education will not weather the storm of AI by tacking its sails to either a total embrace of AI or its complete rejection. AI has unmoored educators in a way that we are still coming to grips with, its implications so immediate, yet so fraught with existential dread. I do know that I am glad that I am no longer fighting this battle from inside the classroom, as I am aware of the fact that I lack the time and energy necessary to completely revise all of my courses in the way they would need to be altered to fit the mold of an AI Enthusiast or AI Resister. That is the task of faculty we have entrusted to convey knowledge and skills to the next generation, even while these students boo graduation speakers at the mere mention of AI (I wonder if they will similarly chastise hiring authorities who ask about their AI competency during post-grad job interviews…).

As a newly minted administrator of a workforce-focused unit, I am now thinking in an entirely different frame—how can I help people in my community gain the skills necessary to find a good job? To me, the answer to this question unquestionably involves emerging technologies like AI. It is, if nothing else, a more straightforward query than the one I faced as a faculty member, albeit one marked by its own perils. That is the challenge I’m excited about tackling, one I feel properly equipped to undertake.

---

Postscript: If you are interested in reading insightful commentary on higher education in the AI age that represents all points along the AI Enthusiast-AI Resister spectrum, I highly recommend following Lily M. Abadal, Tina Austin, Dean W. Ball, Stephen Fitzpatrick, Jason Gulya, Nick Potkalitsky, and Hollis Robbins.