© 2026 WMRA and WEMC
NPR News & NPR Talk 90.7 Central Shenandoah Valley - 103.5 Charlottesville - 89.9 Lexington - 94.5 Winchester - 91.3 Farmville
Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations

The forefront of innovation: JMU Professor Dennis Lo integrates AI into the classroom

JMU students working with AI
JMU students working with AI
Professor Dennis Lo
Professor Dennis Lo

Artificial Intelligence is considered the forefront of innovation, but it is often prohibited in schools. JMU Professor Dennis Lo is integrating AI into his classroom, and provides insight into how it is being used in higher education, the benefits, the dangers, and what it looks like going forwards. WMRA’s Chris Boros speaks with him about AI, starting with his involvement in the JMU AI strategy working group.

* This interview was produced by WMRA's JMU student producer Sydney Sultanik.

Dennis Lo:

So the AI strategy working group Is trying to come up with a coherent plan for orienting JMU towards the next generation of pedagogy and teaching and research in an age of AI. So we've heard a lot about how AI can boost our productivity and efficiency, but the AI working group wants to be more thoughtful and mindful in its approach towards AI. It wants to make sure that we are human-centered, that we are always focusing on the ethics of adopting AI, and that we're making sure that our students and our faculty and our staff are always front and center as we transition into adopting these AI technologies. So the working group is not so much coming up with the nitty gritty policies per se, but more like sketching out a broader path and vision towards the future.

WMRA’s Chris Boros:

And you have a LLM that you use in your classroom, right?

Dennis Lo:

I use AI tools in my classrooms that are based off of a system called RAG. Okay, RAG is called retrieval augmented generation. All right. And that's a really fancy term that basically means the language model can only draw from the sources that you provided. So if you could drag and drop certain files, certain readings, certain texts, it will only answer questions based off of the knowledge in those texts. So it could even answer questions such as "I don't know," right, which is a really powerful answer because we all know about how LLMs are sycophants and they love to actually sound confident, right, when they tell you that "Oh, I know the answer to everything," and then they'll backtrack and say, "Oh, you're absolutely right. I was incorrect," right? But rag models are dependent on what you feed into it. So what I'm drawing, what I'm moving towards, really is this paradigm called critical AI literacy. It's not just being literate about how to use AI; it's being literate about the biases, the ethical problems, right, the sort of issues it might pose for employment in the future or ecological problems that using AI poses, but I want my students to use AI in a responsible and sophisticated manner, so they know how to critique it from within.

WMRA:

Why have you decided to work with AI in your classroom? I mean, is there a reason that you've said yes to AI?

Dennis Lo:

Yeah. So one of the most important reasons is practical.

Right now, a lot of employers are asking for basic critical AI literacy, so they want the students to be familiar with the frontier AI tools out there, but they also want them to know when to use it and when not to use it. And the subject that I teach in film and media studies, right, it's an industry that has very fraught relationships with artificial intelligence. On the one hand, it's extremely afraid that AI will basically take over human labor and digital effects. A lot of work that really takes that artisanal touch, right? And there have been so many walkouts and sort of labor union strikes against AI replacing human creativity, but at the same time, if you talk to Hollywood insiders, they'll talk about how every major film studio has their own frontier-level AI working group that's testing out how to use AI tools to bolster their own production processes. In other words, they have a really complex relationship with AI, and we want our students, right, who go into all these industries, to know the pros and the cons, the short-term benefits, and the long-term sort of consequences of using AI.

The Authorship of Place by Dennis Lo
The Authorship of Place by Dennis Lo

WMRA:

What would be some of those dangerous paths?

Dennis Lo:

You think one of the most dangerous paths is called cognitive offloading. So it's a fancy term for basically saying we are letting AI do all our thinking for us, right? And when you write down a prompt, and you do it often enough, especially with the sycophantic answers that AI gives you, you sometimes come across with this almost narcissistic vibe, right? From using AI that you're somehow all knowledgeable, but the truth is, you are basically just getting results that are calculated off of statistical measurements, and it's not at all reflective of your actual intelligence.And what happens when you actually offload? Right, you give a robot the ability to carry out tasks that are truly based on actual intelligence, actual creativity, you're replacing authentic reality that doesn't have any basis or grounding in truth

WMRA:

. What about your peers at JMU or maybe even at other universities? How are they working with AI? I mean, how many of them are supporting it as you are, or how many of them are like, no, absolutely not.

Dennis Lo:

Yes, I think I'd like to start that with a caveat that I am supporting it to the extent that it is being used in a very thoughtful and careful manner within the prescriptions of a particular class or syllabus, right, or learning outcome. And one of my principles, for instance, is that if you're using AI just for the sake of using AI, if it's not fixing or solving a problem you're noticing from student evaluations, for instance, then there's no need to actually use AI. But that being said, right, there is a whole range of approaches. Just as an example, in the humanities, many of the practitioners, poets, right, painters, filmmakers, there are many creative sort of practitioners and artists who are very much opposed to the use of AI whatsoever. Right, because it feels like this is a genuinely human-centered activity, and that is precisely what you should not be replacing by automation. Automation should replace the more mundane, the more menial chores, not the part that makes making art fun, right? But then there are also you know the particular departments and colleges that need to teach students how to use AI quickly and effectively, so they're able to keep up with how the industry is accelerating, right? And so they don't want to have their own students jump into a software development industry, for instance, with no understanding of AI, and saying in a very condescending manner in a job interview, "How could you ask me, you know, what AI tools have I used? I will not stoop down to that level. I'm a traditionalist. I will code manually by hand." Right?

Somebody like that is not probably going to get hired, even for their kind of attitude, right? So they need to teach them how to use AI tools to accelerate certain skills, but at the same time, still know the basics well enough that they could actually quality control check the results coming out of AI. And as for me, because my realm of teaching and research is in media studies, critical media theory. What I like to do is I like to take the tools that are problematic, and use those tools in a way that goes against the grain of how they're meant to be used, so that I could assess the inherent flaws and biases and problems within those technologies.

WMRA:

What do you hope students take away from using AI in your classrooms?

Dennis Lo:

I think what they can take away is that they understand first and foremost that AI is not a person, it is not a consciousness, right? It is not a being, it is not super intelligence or any of those attributes.

WMRA:

Not yet, right?

Dennis Lo:

Well, yes, there you go. That is the million-dollar question. It really is, isn't it? Yeah. But instead, they understand that it is a tool, right? I won't say it is a tool like any other, right? Because it is such a general tool that it could actually be implemented across so many different mediums and so many different domains. For instance, when you talk about Gen AI, you're not just talking about a chatbot, you're not just talking about video generators. You're talking about algorithms that can teach autonomous cars how to drive. For instance, right? So there's so many different applications of Gen AI, but in the end, there's still technology, right? There's still tools, and we cannot mix them up with actual conscious personas, and as a result, right? I want our students to realize that they can program, they can control, and use these devices in so far as it could further their own critical thinking. I never thought about this before, but I mean, if you aren't encouraging students to use it in some capacity, they are going to be kind of left out of the world that we're looking at in the future. That's right. I wouldn't even use the word encourage. I would say probably expose them to the different kinds of language models, the different kinds of AI tools that are out there, but always giving them the option, right, of saying no to it if they have personal qualms against it as well. A good example is in my classes. I always give them multiple options to complete their assignments.

One good example is from my 160-person large general education class. The students get to actually do the readings. Some very dense, theory-heavy philosophical readings. The traditional style, right? Go through it the way I did back in college and take notes. But you know, after you know several weeks, they may be wondering: should I prioritize the readings, or should I actually be going over the slides, or how should I make the best use of my time? Students are spread so thin these days, right? Now, there's another option, which is you can query a language model that is built off of only those actual texts themselves, right? And you could ask questions about how this theorist thinks about this concept, and how they are in dialogue with another theorist, right?

There's a third approach. You could combine the use of that language model as a reading facilitator, right, to ask these kinds of clarifying questions about terms, but then once you're done clarifying the term, you'll find the ideas really fascinating, and you'll go back into the actual text and read it even more carefully. So what I've discovered is because it becomes much more multidimensional, it's not just reading from the first word all the way to the last one. It becomes much more nonlinear and creative the way the students are mind mapping right across these different forms of reading, and so I call this strategic reading, right? So use AI or not, or hybridize your approaches depending on what allows you to learn best.

WMRA:

What do you hope the future is for artificial intelligence?

Dennis Lo:

Oh, that is a giant question. I would say just to be a little bit more specific in higher education. Sure, let's go there. I would hope that AI could really be seen no longer as a threat, right, to our existence. There's been a lot of you know discussion about how higher ed is in crisis right now, how it's seen as unaffordable, out of touch, right? There are some ideological biases inside it, or that it simply doesn't provide you with the skills needed to survive in an AI-driven accelerated workforce, right? But I feel that what I hope is by allowing people to cultivate this AI critical literacy, they're able to see that actually we don't have to think of AI as the apocalyptic end of the world scenario, because what will survive this transition is the ability to think like a human, to create like a human. So I've been having these you know ideas along with many other colleagues, and I think JMU on the whole is moving in this direction, but we want to think about the university as a human-centered institution, right? It's no longer just a place to teach hard skills.

It's no longer just a place to teach your basic arithmetic, so to speak, right? But instead, it's about teaching people how to do systems level thinking? How to make decisions based on experience in the workplace? How to have actual human social interactions and to understand the nuance of human relationship? All of those things, no matter how sophisticated AI becomes, right? It could simulate those things, but it could never capture something that will allow us to trust its results enough to pretend that it's actually human, right? And so I believe colleges, universities will not only survive but perhaps come out even stronger than before from this transition by focusing even more on the human aspect of our learning. Streamline really what is taught in the classroom, not just rote knowledge, not just skills, but the stuff that allows us to be having this conversation. Dennis Lowe is an associate professor of global cinemas in the English Department at James Madison University and a member of JMU's AI Strategy Working Group.

WMRA:

Dennis, great talking to you today. Thanks so much.

Dennis Lo:

It's wonderful talking with you too, Chris. Thank you so much for inviting me.

Sydney Sultanik is WMRA's JMU Student Producer
Chris Boros is WMRA’s Program Director and local host from 10am-4pm Monday-Friday.