Andy (00:01)
Welcome to the HigherEdJobs Podcast. I'm Andy Hibel, the chief operating officer and one of the co-founders of HigherEdJobs.

Kelly (00:07)
And I'm Kelly Cherwin, the Director of Editorial Strategy. Today we're joined by Dr. José Antonio Bowen and Dr. Edward Watson, two national leaders helping campuses navigate rapid changes in teaching and learning where AI is involved. Dr. Bowen is a former university president, dean, and award-winning educator known for driving innovation in the classroom. And Dr. Watson is the vice president for digital innovation at AAC&U, the American Association of Colleges and Universities, and a leading voice on AI, digital equity, and the future of undergraduate learning.

Andy (00:38)
They've released the second edition of their book, *Teaching with AI: A Practical Guide to a New Era of Human Learning*, which spans everything from ethics and privacy to custom bots, individualized feedback, and the realities of how students are already using AI.

Mike (00:54)
Welcome back. In the first half of our conversation, we explored the grief and anxiety many faculty members are feeling as AI challenges their existential identity as subject matter experts. We were left with a powerful warning: if we simply cognitively offload our creativity to these tools, we risk damaging human thinking significantly. Let's rejoin the conversation as we shift focus toward the potential of AI to become a true partner in thinking, helping create and design better outcomes than what a human or bot could produce individually.

Eddie (01:30)
So, you know, the medical field offers two different views. I mean, one of the things that José and I often say is, you know, what we call cheating in higher education, business is calling progress. So how do we navigate a landscape where what we may see as cheating might actually be a skill that students need to possess post-graduation? So really, it's a complex space that I think the learning outcome and really the professor's own preference would drive—what the choices might be in an individual class regarding the use of AI.

Jose (02:03)
And there's a new study about doctors that came out this week. Previously, note-taking has been something that 66 or 67 percent of doctors were using. Now doctors are using it daily. So note-taking is easy, insurance codes. But 80-some percent of doctors said it's making them better doctors. Some of that's because they can listen better to patients. But a new use for doctors is getting alternative diagnoses.

Which is interesting, right? Because think about that. So here's what I think it is: what else might I be missing? Right? This is exactly like how we would use it to make our writing better. So clearly, for a medical student, you don't want to short-circuit that, right? You want to think about all the possible diagnoses yourself. But if you're an expert—you're already a doctor and you already think you know what the diagnosis is—having a partner who says, well, it could be this.

And so this remarkable statistic that 84 percent of doctors or something like that think that AI is making them better doctors is, I think, really important for us to realize that the question is not, do we need to introduce AI? It's when and how. It's like the calculator. It might not be the first thing we have you do. We might have you learn to do math the hard way, which actually we should have, right? But at some point, we've got to introduce the calculator and the spreadsheet and other kinds of tools.

And so that's a pedagogical question, and all of us need to think about that. Where do we want students to use AI, and where do we not? And why? As Eddie says, what are the learning outcomes? And that'll be different if I'm writing a lab report versus learning how to write.

Andy (03:42)
Wow, that's so interesting. The medical example, which I'll come back to in just a sec, is really interesting on the human side of things. But also with that last reflection, José, is the point where we feel like academia finally adopted AI wholeheartedly when we acknowledge that the most unbiased peer reviewer is going to be AI, if we're going in that direction.

Or maybe certain peer review organizations are going to want some bias within their AI. So I want a peer reviewer who has a chip on their shoulder, and now AI is going to be able to do that exactly the way somebody with a chip on their shoulder does. That really kind of piqued my interest, particularly within the internal structure of academia.

But on the human side, with the doctors and the nurses and the kind of disparate use there, it really kind of edges on this, but doesn't quite hit it perfectly. And I think this is one of the concerns that you see repeatedly throughout society about AI. The book does a good job of expanding on AI-supported tutoring, feedback, and customized learning. And that's exactly what we're talking about with doctors and nurses.

But what are the most promising ways that AI can individualize learning without weakening the ability to develop crucial human relationships that we could possibly, like you said in these other instances, start skipping if we have AI? Human-to-human contact is just so important.

And it's a bad example, but just watching it last night with my wife—the Apple show *Pluribus*, which is not completely finished yet at the time we're recording this—the idea of human contact in that show is definitely a question. No spoiler alerts here; I'm not going to spoil *Pluribus* for the entire academic world. But that part of AI, and maybe I'm speaking a little bit for myself, when you really sit with it sometimes feels like the most scary. What do you think are the most promising ways that AI is going to allow us to grow the way it should, but not weaken our ability to interact with others?

Jose (06:01)
Well, there are a couple of questions in there. The first is, I don't think we know the answer to your last question. Let me start by saying I believe that human relationships are what drive a great education. And I think there's some evidence for that. But I also know that my belief that AI can't replace those relationships is a belief at this point, right? That the customized AI tutor, for some people, might be able to—I don't know.

I want the relationship part to be true, and I believe in the evidence that we have so far, but we're entering a new age. And just in the same way that we used to not believe that you could work from home, that you had to be in the office. And so old people like me, right, want people to be in the office, right? But young people are like, nah, I get more work done. And the evidence says that you get more work done at home. There are negative things that happen, so I think it's different for different people.

I think we don't yet know, so we need to pay attention to where are the places where relationships really matter. Right? We often think, advising—I need to be in the room with the student. Well, you do. But a lot of advising is not advising; it's administration. It's: did you take the prerequisite? You know, does this schedule work for you? Hey, AI could do that piece. The relationship piece is: how are you feeling? How do you feel about your education here?

And what we often do is we have an administrative conversation about prerequisites and when you're going to study abroad and think, well, I had a relationship, right? And so where is the place where relationships really matter for students? So I do think we need to keep an open mind about where this could be.

But to your other question, maybe I'll let Eddie start with where we use customization and the other things that AI can do.

Eddie (07:56)
Well, you know, I guess I'm sort of following up on some of your initial thoughts, José. I mean, I have some dystopian views of what the future might look like. And if we could predict the future, I think we would probably not be on this podcast—we'd be doing other kinds of things right now.

But one possible utopian future for higher education and AI—I think this is likely possible—we will see this capitalized upon in some very specific pockets. But, you know, there's currently great resistance from faculty to the notion of AI serving as a grader within a course. Like the recent AAC&U faculty survey really highlights that the vast majority of faculty feel like that's an act of academic cheating.

But as we think about the variety of ways that AI could be leveraged for a number of the administrative components of the process of teaching—you know, there's the time that you spend in the classroom, but there's also class prep, assignment design, grading, a number of different things—I guess I'm hopeful that maybe a new signature pedagogy could emerge for higher education across all disciplines.

That as we might capitalize upon improving generative AI tools that can help us with some of the administrative tasks, instead of maybe teaching a four-and-four load equaling 40 hours a week—roughly, give or take, maybe that's a rosy notion—but if you leveraged AI in a way that instead of spending 40 hours a week prepping and grading and everything that we do for teaching, maybe it would be more like 35 hours. Maybe it's 32 hours if the tools improved and we trusted them.

Well, then what might we do with that extra time? Might we reinvest in our students, in those personal relationships? Where, you know, if we know that we've got an extra eight hours every week, how might we make ourselves more available to our students? Of course, students have to meet us there, right? But, you know, to say, hey, I'm going to be in the dining hall from 12 to 2 every Tuesday and Thursday, and I'd love to talk about my career journey or what the job possibilities are, or the thing that we're going to be talking about in class tomorrow. But come and talk with me about anything.

We can talk about surfing. We can talk about running. We can talk about your career, your aspirations, developing relationships—all of those kinds of things that it's really kind of hard to do these days, you know, when you've got that many students coming in through a four-and-four teaching load. So it's possible that actually AI itself isn't the element that increases personalization, but rather it frees up time so that the human spends more time with human students to engage in personalization and making connections in their lives.

Jose (10:51)
And here's another example of how AI could help us. So most of us give generic assignments, even though we know that math students don't really care about Train A leaving the station going 20 miles an hour faster than Train B. We know that unless you want to be a train scheduler, nobody really cares, right?

So we know that it's much more effective—and students will actually try harder and put more effort in—if the question is, well, the wide receiver you're covering runs 10 miles faster than you do, or the Republicans are registering 30 voters an hour and the Democrats are registering 20 voters an hour. How much earlier did that happen, right? That's exactly the same math problem.

And so most of us could say, okay, I'm going to do three versions and you take your pick, but it's the same math. But AI now allows us to customize every single problem set for every single student. So I could actually give you problems that relate to your interests. You want to be an anthropology major, you're a poli sci major, you're a business major, or you're interested in football.

So we could give every student a problem set that's more motivating. Now it's not going to take us any time. That used to be a real issue—it was too time-consuming; we couldn't do it. So finding the places where AI could help, as Eddie suggests, is I think the first step.

But to come back to your question, Andy, ultimately you're asking a social question, right? Do we accept that people can work from home now or be on a Zoom call instead of having to have a real meeting? And that we changed what we assumed was okay.

So at the moment, we don't want our doctor to go in the hallway, look up AI, and say, here's what you have, and come back in. On the other hand, we're fine with our doctor stepping into the hallway and not doing long division, which was the case when I was a kid, right? The doctor had to do long division on a piece of paper to know how many milligrams to give you. And now the doctor uses an app, and we accept that as being better and more accurate.

So I want the doctor's opinion, but I also might want a second opinion from AI, or I might want the doctor to look at other possibilities suggested by AI. Some of that is using the expertise that the doctor has, but some of that is also social.

It's a little bit like self-driving cars, right? We're nervous about self-driving cars, even though they're already significantly safer than human-driven cars, right? So the problem with self-driving cars is really our human anxiety. And so at some point, could we get to the place where we ban humans from driving altogether? We would save hundreds of lives a day—a day—but it feels weird.

And so there's a technology question, but there's also just a question of what we feel comfortable with. And I think we're in that place for all of us, not just faculty, where there's anxiety, there's grief, there's identity. And so it's very hard to know if in five years everybody will accept AI plus human as the writing standard and just assume that every piece of writing you see has an AI component. I don't know.

Andy (13:57)
When do you think *The Wall Street Journal* and *The New York Times* will start using spell check?

Jose (14:02)
[laughs]

Andy (14:03)
But that's kind of like—it’s one of those places where, when you talk about doctors or all these examples, little by little we've had things creep into our society where those tools are there. My gosh, I remember a point in my academic training when you couldn't use a calculator in a math class. Now there are so many different places where you can use a calculator. But boy, I can tell you there were points where I was not allowed to use one.

I know we could go on forever, but I want to make sure—I've already plugged *Pluribus*—I want to definitely wrap up here by saying I'd love to leave our listeners with some concrete next steps. Please read the book. It's fantastic. But are there any new chapters in the book that you'd like to highlight?

Jose (14:50)
So there's a new chapter in the book on custom bots. That's not a technology that was really developed when the first edition came out. So we've expanded the sections on role-playing and feedback, but especially on how to design a custom bot to help your students. You could design a syllabus—a course tutor, essentially, using the syllabus. You could do an assignment bot, or you could just create a bot that replaces an exam.

So instead of taking the midterm, you talk to Einstein, the bot, who then determines when you've learned enough thermodynamics and will actually help you get to that point. So there's a whole new chapter on that. There's also the chapter on literacy, which we sort of talked about earlier. So that's the other brand-new chapter.

Eddie (15:35)
I'll add a little bit about that. So our broadly updated chapter on AI literacy actually surveyed the landscape of the variety of different AI literacy frameworks and models that have emerged and looked at where there were commonalities across those models.

I would not say that we developed a new framework, but I would say that it's a meta-analysis or a meta-framework of the various components that you see out there in the landscape. And we added some additional ideas to that as well that we hope will serve as a guide for higher education, at least for the next year or so, until maybe our third edition of the book comes out.

Andy (16:18)
The book publisher might need to write the book in pencil at this point. In addition to reading the book—assuming everybody who's listening is going to—are there other resources or tools that the two of you would recommend for faculty who want to continue learning and experimenting in this space?

Jose (16:37)
So we do have a website, We Teach with AI, and it has all the prompts from the book, but it also has an up-to-date list of models. There are many more models because they're coming out all the time. There are 30 different linguistic, regional, and cultural models, for example, listed there.

There's also a list of resources, which includes things like university policies and faculty development. There's prompting advice, there's assignment ideas. And so send us assignment ideas and we add them to the website. It is a collective project, but that's a place where people can go, click on some links, try different models, and get ideas for new assignments.

Andy (17:17)
And can we just put the website out there again?

Jose (17:19)
Yes—weteachwithai.com—and we'll put that in your show notes or link to it someplace.

Eddie (17:26)
Yeah, and I would share that AAC&U—we are leaning into a wide array of programming to support the entire institution. I mean, José and I do a Teaching with AI workshop series every semester. The forthcoming ones in the month of February will be doing something similar, certainly updated for the fall semester of 2026 and beyond.

We have an institute on AI pedagogy and the curriculum where campuses send teams to engage in curricular revision work and pedagogical change and policy development and governance—all those kinds of things. We have a symposium on AI leadership for those who are trying to lead change on their campus.

I mean, we've developed programming for sort of every strata of higher education to help institutions thoughtfully engage in thinking about their next steps as a result of the emergence of generative AI and agentic AI and whatever comes after that.

Kelly (18:25)
Thank you for all those fantastic resources. Obviously, if the listener hasn't gotten the point, please read this book. It's fantastic. José, Eddie, it's been such a fun conversation. It's been great to have you on today. Thank you for joining us.

Jose (18:37)
Thank you.

Andy (18:38)
And thank you for listening. We hope you enjoyed the conversation as much as we did. If you have any questions for José or Eddie, please feel free to email us at [podcast@HigherEdJobs.com](mailto:podcast@HigherEdJobs.com) or send us a direct message on X @higheredcareers. Any thoughts—anything—please. We want to hear from you. Thank you, and we look forward to talking with you all soon.