"...most people complain about universities forcing AI on people, but that’s generally not the case."

Part 2 of an interview with Bryan Alexander

In the last post, we talked with Bryan Alexander about his own AI usage as well as began discussing his new book, Peak Higher Ed: How to Survive the Looming Academic Crisis. We continue that conversation here.

---

About the Interview series

As part of my work in this space, I seek to highlight the folks I’ve been in conversation with or learning from over the last few years as we navigate teaching and learning in the age of AI.

If you have experiences around AI and education, particularly in the higher education space, that you would like to share, consider being interviewed for this Substack. Whether it is a voice of concern, curiosity, or confusion, we need you all to be part of the conversation!

---

Eaton: I appreciate that distinction between web and internet. Let’s look at your discipline scenarios, when you talk about this potential of two coexisting groups: those who lean into AI and those who lean away from it, anti-AI or otherwise. I understand this is hypothetical but through your lens, how much of that do you think will end up being class-based?

I’m thinking here about who gets to opt in and opt out, and who’s expected to. Virginia Eubanks’s Automating Inequality comes to mind. How and where do these systems show up? In your discipline scenarios, do you have thoughts about who is opted in or opted out, or anything along those lines?

Alexander: For the discipline scenario, I’m drawing on the work of the great futurist James Dator from the University of Hawaiʻi. He came up with four different archetypes for how we think about the future and how the future of a system, a world, or a business might unfold. One of those archetypes is discipline.

By discipline, he means a massive reorganization around controlling something. For me, the classic examples are strong religious reorganizations—not saying good or bad—like the Reformation in Europe. We might think about Afghanistan. For many futurists, this shows up in discussions of consumerism: what if a society moves toward degrowth or voluntary simplicity?

Here, I think about controlling AI. We actually have plenty of precedents for this. People often talk as if technology inevitably runs wild, but that’s not always the case. We regulate lots of technologies, and some we clamp down on AI hard.

Go back to the 1940s and 1950s and you’ll see people going completely bananas over atomic power. It is now massively regulated and, in fact, cut back quite a bit in the U.S. and other countries. Think about biotechnology, which is incredibly regulated and tightly controlled. Or think about how hard it is to build an atomic bomb in your backyard. I’m glad that’s true.

It may be that we come to view AI this way. Again, this is a question of perception. That it’s something threatening, unreliable, unstable, or dangerous, and therefore something we want to corral and control. There are all kinds of obstacles and issues around that, but that’s one possible path.

It may also be that this happens from the bottom up. People proclaim, “I’m refusing AI.” I’ve seen this with some digital creators: “My content is 100 percent human-generated.” Others will say, “Mine is 70 percent human-generated—here are the AI bits.” And then you have people who just glory in it, who say, “This is amazing what I’m able to do.” You have people who make short videos and turn them into twenty-minute short films. We might see that kind of divide play out.

In academia, we’re already starting to see this. One of the things that fascinates me about academic responses to AI is that a big majority, about 76 percent, of colleges and universities still don’t have strategic policies about AI. I’ve been tracking this steadily since it emerged. Most of the policies you actually see are basically the academic integrity policy: the “do not cheat” policy with “don’t do it with AI” stuck into the middle of it.

And in fact, most people complain about universities forcing AI on people, but that’s generally not the case. Overall, universities are waiting to see how this is going to go. They push decision-making down from the strategic level to academic officers, who then defer it. Sometimes they’ll pay for professional development, which is great, because we need a lot more of it. But mostly they defer to units: schools, colleges, departments, which then defer to individual departments, and eventually it trickles down to individual faculty making decisions.

Keep in mind, this is happening in an atmosphere of serious financial stress. Being able to afford massive professional development and institutional rethinking is not something a lot of campuses can do. When this devolves to individual faculty, we get this fascinating situation that I don’t think people fully realize: the largest proportion of faculty in the United States are not tenure-track. They are adjuncts: people who are paid horribly, who usually have no institutional support in terms of offices or staff, much less healthcare or retirement.

And those people are, in effect, leading American academia’s response to AI, which I find fascinating. I’ve had some very strange conversations about this. But given that distributed decision-making process—not really bottom-up, but very dispersed—we’ve already seen individual faculty make very different calls.

I just saw one professor railing against AI as awful and explicitly saying, “You may not use AI in this class. You have to write to me if you want an exception.” And then you have other professors saying, “We are going to use AI in this class.” Others have no policy at all.

We may see that division play out more and more. And this is where students really run into problems. If you’re taking a French class where the professor says you may use AI, and then you go to your math class where the professor says you may not, you start having to navigate where you can and can’t use it. If you use AI in your life more generally, you have to carve out routes and exceptions. That’s very, very challenging.

And that’s not even counting the difference between what faculty decide to do in their classes and what students do in the rest of their lives. I’ve talked to many professors who have banned all digital devices from their classrooms as a result of AI. They think, “Okay, this way I can separate my students from AI.” But the minute students walk out the classroom door, out comes the phone or tablet or laptop, and they’re merrily back on Gemini or Claude again.

We may see this discipline scenario take shape in various forms. We may see more writing programs advocate for no or low AI across composition, especially writing across the curriculum. We may see other disciplines push for more AI such as economics and computer science. We may eventually see colleges and universities start making major, institution-wide decisions about this. But that kind of discipline is one thing we really have to anticipate.

Eaton: And let me push a little further on that, because when I’m thinking about this—and this ties into adjunctification—adjunctification also happens at a larger scale at certain institutions than others. I’m thinking about massive online institutions, right? Those are almost entirely run by part-time faculty.

My guess is those are also—at least from what I’ve started to see—the first places that are using AI in different parts of the system more actively. And what I’m trying to question is how the discipline approach might also help us think about who gets to opt into AI and who can’t really make that decision, or feels like they can’t say no.

Alexander: Yeah, there are material constraints here.

One thing to think about: we started to see this emerge in the early 2000s: the idea that some wealthy families would deliberately deemphasize digital tools for their kids. That itself became a sign of wealth and class. At the same time, we saw more digital tools being deployed in lower-income contexts. And we’ve seen this play out in different ways.

Chris Gilliard has done great work showing how technologies of surveillance are widespread, but are often more heavily aimed at lower-income people, whereas very wealthy people tend to have more options for escaping that kind of monitoring. We could see something similar play out here.

If you think about access to postsecondary education, who feels they can afford it, who feels prepared for it, Robert Putnam documents this very carefully in Our Kids. Despite colleges and universities having all kinds of aid programs available, a lot of students from poorer communities never learned about college preparation in middle school or high school. Even if they could apply, even if they had a chance of getting in with a good financial aid package, they often didn’t have the preparation. They didn’t have guidance counselors. They didn’t have teachers encouraging them to pursue those paths.

We might see a similar divide emerge around AI, especially in financially stressed institutions and communities.

Right now, there are two big forces holding things back. One is fear. This sense that replacing educators with AI is frightening. That’s a classic problem. We have two hundred years of industrial revolutions wrestling with what it means to replace human labor with technology. It’s always anxious and fraught, involving labor politics and public attitudes. We may even decide, culturally, that we love teachers and don’t want to see them replaced by machines.

The other force is financial pressure. These tools are expensive to deploy at the enterprise level. As individuals, we can fire up ChatGPT and use it for free or very cheaply. But for a school system, it’s horrendously expensive.

The California State University system struck a very good deal financially with OpenAI. Otherwise, every CIO I talk to says this is awful: they don’t know what to do because the tools are so expensive, especially if you want anything that’s actually good. So it may be that once prices come down, or if more schools decide to use open-source tools, we’ll see broader adoption.

Eaton: My own working theory is that there was a time when giving all faculty, staff, and students access to the Microsoft Office suite was unheard of, and now it’s just a given. There’s something similar here around finding the price point and planning for it. We saw the same thing with institutions developing full access to the internet on campus; it became planned infrastructure. But at that point, I do wonder what it will mean for choice for people.

Shifting gears a bit to that question of resistance around AI in academia. We’re seeing certain disciplines such as writing (and rightfully so), vocally concerned about these tools. Do you see significant variation across disciplines in their interest in or resistance to AI? Are there some fields that are just structurally positioned differently to respond?

Alexander: Sure. It’s two cultures of C. P. Snow’s Two Cultures. The opposition to AI that I’m seeing comes from the humanities. I’m seeing it from American studies scholars like Matt Seybold, from literature scholars, philosophers, historians: people in writing, rhetoric, and composition. Overwhelmingly, that’s where it’s coming from right now. And it draws on a long history of humanists feeling uncomfortable with, or opposed to, technology.

You can hear it when you talk to people. They’ll say, “All right, this is the last straw. I wasn’t happy about laptops in my classroom, but now we’ve gone too far.”

The social sciences are interesting. The more quantitatively intensive ones tend to line up closer to the sciences, and they’re more likely to say, “This is a tool we’re using.” But the non-quantitative social sciences—parts of sociology or psychology that don’t rely heavily on statistics—I’m not sure where they are yet.

But I really do think this is C. P. Snow’s world; we just live in it. And it makes sense for a lot of reasons. You mentioned writing, but there’s also a strong political and institutional layer here. One argument I hear from a lot of humanists is that they feel under attack by their administrations. They feel defunded, deemphasized, and pushed aside in favor of STEM. They think this is a mistake: wrong for universities and colleges, and frankly inhumane. And they see AI as something being imposed on them by administrators.

There’s also a political dimension. Some people align AI with the right, thinking about Republican support for it, or Silicon Valley leaders aligning with the Trump administration. Others align it with capitalism from a different angle, seeing it as late capitalism—although that phrase itself is about a hundred years old—another instance of markets replacing labor with capital and mechanization. That political critique cuts across disciplines, but the disciplinary divide itself is pretty clear.

Eaton: Thinking about the chapter on AI as a whole, and that idea of looking for signals and doing horizon scanning: what kinds of signals should institutional leaders be watching to distinguish between these scenarios in real time? At what point should strategic planning shift from anticipating one future to another?

Alexander: I have a very unpopular bit of advice on this, which is that campuses need to mobilize around AI the way we did during COVID. We need a continuous intelligence-gathering operation.

And I don’t mean, “Fred in computer science can tell us what’s going on.” I mean a genuinely interdisciplinary, diverse effort across multiple axes. It should be ongoing and constantly sharing within the campus what’s happening with AI. It should involve IT, the library, instructional designers, people from all academic disciplines, the social sciences, the humanities, the arts. It should absolutely involve students, as well as staff and faculty.

During COVID, lots of universities created exactly this kind of committee or team. We need that again, because AI is vast, complex, and evolving very quickly. Without that kind of collective intelligence-gathering, campuses are flying blind.

A second angle is doing a deep dive into an institution’s own use of AI. Right now, it’s very hard to get good data on actual use. If you poll people, you run into the same problems we see when polling about math anxiety, sex, or drug use: people have all kinds of reasons to lie.

If you ask a student, “Do you use ChatGPT or generative AI?” and they think they’ve been told it’s sketchy or untrustworthy, they may say no, even if they do use it. Or, if they think using AI signals that they’re savvy and workplace-ready, they may overstate their use. It’s even worse when you ask faculty or staff.

I see a lot of faculty and staff using chatbots for fairly basic but consequential tasks: writing grant proposals, drafting RFPs, digesting complicated documents, handling email, munging data. But we don’t have good surveys capturing that. It would be incredibly valuable for individual campuses to develop ways of gathering and sharing this information locally, so people can say, “Here’s what’s actually happening. Here’s how people are using these tools.”

Once you have that, institutions should be watching a whole set of signals. We should be paying attention to the financial health of generative AI firms. Are we in a bubble, or heading toward a market correction? We should be tracking pedagogical models, from Niall Ferguson’s The Cloister and The Starship framing to the many different ways people are teaching writing with AI, and asking: do these benefits outweigh the costs? Are they actually helping students?

And we have to look carefully at the labor market. I know whenever I say this, people respond, “Higher education isn’t job training.” But preparing students for the labor market in multiple ways is a core part of postsecondary education. Ignoring that is unrealistic. Students are thinking about this intensely, not because they’re brainwashed, but because they’re facing a difficult economy and real debt burdens.

We need to examine what AI is doing to the labor market. Where are we seeing actual job replacement? That’s still unclear. Where is AI being infused into everyday workplace tasks? That’s harder to track, but economists can study this, and we should be drawing on that expertise. Then, based on all of that, institutions need to pivot their strategies accordingly.

Those are some of the key things leaders should be watching.

Eaton: Final question. There’s always a bit of time between when you finish a book and when it’s published. Have there been other insights that have emerged since then, either about the role of AI in higher education or about peak higher ed more broadly?

Alexander: Whenever you write a book, as opposed to making digital content, there’s always a lag time, and I’m always worried about that. One thing that’s really stood out since the book went to press is that the second Trump administration’s attack on higher education has been even more severe than I expected.

To back up: over the past couple of years, I did a lot of work around this. I ran a large open online reading of Project 2025. I did a series of Future Trends Forum sessions where we tried to analyze and anticipate what a Trump administration 2.0 would look like. We got a lot right, but some developments have gone even further than we anticipated. A lot of these pressures are accelerating the downward forces on American higher education.

If I could go back in time, this is something I would have emphasized even more in the book.

A second insight is about how higher education is transforming structurally. The National Student Clearinghouse Research Center released a new report estimating that total enrollment has bounced back to roughly 2019 levels. But the shape of higher education has changed.

They point out a few key things. The biggest sources of enrollment growth were not liberal arts colleges and not R1 universities. Instead, they were community colleges, primarily driven by dual enrollment where community colleges are teaching high school students. Growth was also driven by certificates and short-term credentials. It’s not students flocking to PhDs or even master’s degrees; master’s enrollments are beginning to decline. It’s not even primarily bachelor’s degrees. It’s associate degrees and credentials below that level.

We’re seeing higher education morph toward job preparation, and we’re seeing students, families, and their supporters emphasizing career preparation to a degree we haven’t seen before. That’s a qualitative transformation. I talk about it in the book, but I wish I had emphasized it more as an outcome of the quantitative transformation.

Those are two big things. A third is online teaching. I don’t talk about that much in this book. I talked about it extensively in my previous two books, but online education has absolutely boomed. It’s enormous. I don’t think people fully appreciate how much online classes are now a mainstay of American higher education. And how AI intersects with that—how it reshapes it—is still a wide-open question.

Eaton: I feel like that’s a cue for agentic AI and another conversation that might be thrice as long!

---

The Update Space

Upcoming Sightings & Shenanigans

• Writing with AI: Making Informed, Defensible Choices Before You Submit. Free Webinar with Paperpal. Wednesday, April 22 at 3pm (ET).

• EDUCAUSE Online Program: Teaching with AI. Virtual. Facilitating sessions: ongoing

Recent Recordings, Resources, & Writings:

• Damm, C., & Eaton, L. (2026, March). From prompt to practice: A framework for transparent GenAI use in higher education. EDUCAUSE Review.

• Eaton, L., Nemeroff, A., & Sun, X. (2026). AI-assisted course design and development. In K. S. Ives, M. Cini, & R. Schroeder (Eds.), AI applications in online higher education administration: Strategies for maximizing returns and improving outcomes. Routledge.

• Margin of Thought with Priten: Season 1, Episode 5: How Can We Center Pedagogy During the AI Tech Wave? (February 2026)

• Online Learning in the Second Half with John Nash and Jason Johnston: EP 39 - The Higher Ed AI Solution: Good Pedagogy (January 2026)

• The Peer Review Podcast with Sarah Bunin Benor and Mira Sucharov: Authentic Assessment: Co-Creating AI Policies with Students (December 2025)

• David Bachman interviewed me on his Substack, Entropy Bonus (November 2025)

• The AI Diatribe Podcast with Jason Low (November): Episode 17: Can Universities Keep Pace With AI?

• The Opposite of Cheating Podcast with Dr. Tricia Bertram Gallant (October 2025): Season 2, Episode 31.

• The Learning Stack Podcast with Thomas Thompson (August 2025). “(i)nnovations, AI, Pirates, and Access”.

• Intentional Teaching Podcast with Derek Bruff (August 2025). Episode 73: Study Hall with Lance Eaton, Michelle D. Miller, and David Nelson.

• Dissertation: Elbow Patches To Eye Patches: A Phenomenographic Study Of Scholarly Practices, Research Literature Access, And Academic Piracy

• AI Syllabi Policy Repository: 200+ policies (always looking for more- submit your AI syllabus policy here)

Finally, if you are doing interesting things with AI in the teaching and learning space, particularly for higher education, consider being interviewed for this Substack or even contributing. Complete this form, and I’ll get back to you soon!

We periodically host small-group workshops and leadership sessions for higher ed teams. You can learn more about our current offerings here.

---

AI+Edu=Simplified by Lance Eaton is licensed under Attribution-ShareAlike 4.0 International