I spoke with Nathaniel Myers, an associate teaching professor in the University Writing Program whose research interests include digital literacy, generative artificial intelligence, and writing, to get both his opinion and experience in working with students and AI.
I wrote interview questions for Myers, and then, for my own edification and experience, I wrote a prompt for questions using Gemini, Google’s generative AI model. The AI-generated questions were meant as a brainstorming tool and were incorporated into the overall questions for this article. Additionally, the transcript from the recorded interview was transcribed by AI using Rev transcription services. I reviewed the transcript, cleaning it up for accuracy. I edited the article for clarity, conciseness, and space. Lastly, the story received one final edit by our copy editor.
Miller: Tell us about your position.
Myers: I teach first-year students who are primarily required to take the Writing and Rhetoric course that I teach and that my colleagues teach in the writing program. And that course is largely about argumentative kinds of writing and research methods and things like that.
Miller: So what is your involvement with AI?
Myers: My involvement with AI began two years ago when ChatGPT hit, and all of a sudden we had a tool that was able to write a pretty good student essay—maybe not A-level, maybe not even B-level, but at the time, maybe a pretty average C-level paper. I would say that a lot of faculty, especially writing faculty and faculty who teach writing assignments, became very concerned about what they were doing in the class and making sure that students were recognizing the value of things like writing for developing their knowledge and for learning in ways where they weren’t just circumventing that work to AI.
The initial drive was to sort what we are doing now as faculty teaching writing that can be responsive toward students, while recognizing ways in which students might use tools like ChatGPT in our classes. Ways that don’t just have them cognitively offloading the important work of writing, but might actually support and help them grow as learners and writers.
In the past couple of years I’ve worked with Notre Dame Learning to help support faculty growth around AI literacy and responsible AI use.
Miller: What is the AI literacy initiative?
Myers: So AI literacy, or critical AI literacy, is the idea that you have an awareness of generally how AI works, or at least how generative AI works. I think part of AI literacy is knowing that there is AI and then there’s generative AI, and those are related but distinct things. ChatGPT is generative AI. It generates language using predictive analysis or predictive responses based on analysis of large bodies of data and language that it’s pulling from and that it’s trained off of.
But that’s distinct from AI or machine learning, which functions in a slightly different way from those statistical models. AI literacy is about understanding general principles about how it works, but also being aware of how you might use it to engineer prompts to create better responses and more helpful responses from AI. And also to recognize how these tools can create misinformation, and have biases within them for various reasons.
Miller: What are some of the potential ethical concerns around AI?
Myers: There’s an environmental impact of AI, and you want to be thoughtful about the ways in which these companies might be using your data. There are potential data privacy concerns. For example, you don’t want to supply FERPA or HIPAA information into a question that you asked ChatGPT, because they might use that data and then actually present it to somebody else because it’s informing their training. There’s also the awareness of the fact that these tools feel like they’re human, but they’re not human.
Miller: Do you think AI might encourage plagiarism? What if a person prompted Gemini to write a 500-word story—would that be considered plagiarism?
Myers: Well, I think we’re still figuring that out, but I would say that is one of the essential concerns that my colleagues across the University, especially those who teach writing assignments, are very concerned about. I think the idea of academic integrity or academic dishonesty is front and center as a result of the release of ChatGPT. So, to your point, I would hope that we as faculty generally have trust and faith in our students that they want to engage meaningfully with the work that we do with them, but we also recognize that they have busy lives and sometimes they have an assignment due in two hours. And when you’re in that situation, then you might turn to something like ChatGPT to help assist in that work. And is that plagiaristic?
Let’s say I turn to generative AI when I’ve got a paper due. I’m thinking about writing this on such-and-such topic. Can you use the tool to think through some ways into that topic? Well, if you’re not simply presenting the ideas or writing of ChatGPT as your own, but it’s presenting ideas that help you think through the thing that you want to write, is that any different than if you, say, turn to your sibling or your roommate and say, “Hey, I’ve got this paper to write soon. This is my topic. I’m trying to think through it. Can you help me think through this?” I think that’s a harder question to answer. And so how can we use the tools to help without taking away from developing student learning? I think we’re still figuring that out. And we are trying to help faculty figure that out.
Miller: What are your initial thoughts on the potential of AI to impact the work of University faculty and staff?
Myers: I think we all have to reevaluate what it is we are asking our students to do and what are the best practices and methods to do that work. Essentially, I think it’s a question about the learning goals or objectives of your course. And I ask, where do those learning goals maybe need to change because students now have generative AI to assist in that work? And where is it necessary for learning goals to say the same? Because what is important is that students are doing a certain kind of thinking on their own without the assistance of generative AI. But then, how do you think about your policies and your assignment design and your assessment practices that disincentivize students’ desire to turn to generative AI when you do have assignments that require them to use those AI tools, or, alternatively, to think for themselves and not use those tools?
Miller: What are some specific areas within the University where you believe AI could have the most significant positive impact?
Myers: One thing that I have found very helpful for me to think through is brainstorming active learning exercises in the classroom. Let’s say I want to talk about thesis statements with students and I get up at the front of the class and lecture students for 45 minutes on theses, which I don’t think is necessarily the best approach for student learning. If I were to turn to generative AI and say, “Hey, can you help me think through what are some possible active learning exercises around theses that get students more actively engaged”—so they’re not just listening to me talk, but doing some kind of game or some kind of group discussion or something like that that helps them achieve the same learning goals that I want, but do it in a way that’s more engaging for them.
I think the other thing is to think about how you can use AI and bring it into the class for student assignments that allow students to still get to the same knowledge or growth using AI as a way to do some of that work.
For example, last fall, I taught a course called Writing in the Age of AI, where it was an advanced writing and rhetoric course where students openly engaged AI to think about where is it valuable in our writing, and where is it not valuable?
One of the assignments was asking students to use NotebookLM (a Google AI program) as a kind of research assistant to assist in the work of comprehending and working with secondary sources for a research paper they were writing. Some sources can be very jargon-heavy, dense, and hard to get through. We used NotebookLM, which does the work of synthesizing some of those more dense papers, to aid in comprehension.
But then I also asked them to read the works themselves and think critically about what NotebookLM did not capture that was important in their work. It actually became a way to help foster critical thinking toward those secondary sources, toward AI in particular, as a tool, and what were the limitations of the tool itself. It was still getting at the work of critical
thinking and reading and reading comprehension.
Miller: Did you witness students learning from that exercise?
Myers: That class had reflection-based assessments. I had them write responses to their experiences. I didn’t know, for example, what they would walk away with in having them read the actual paper versus what NotebookLM presented to them. It was the students themselves who said in their reflection, “What I thought was the important part of this research paper, NotebookLM didn’t capture.” And so yes, I was able to see that in their reflection work. They helped me observe the limitations of those tools.
Miller: How can the University ensure that AI tools are accessible and equitable for all members of the community?
Myers: I think that’s a real concern. From a teaching perspective, I would say that one of the possibilities of generative AI is that it can possibly help students who may be neurodivergent or who may come from different educational backgrounds to create greater access to the things that are being learned in the classroom.
There’s also the idea that if you have a multilingual student or English as a second language learner, AI might help them draft a paper from their own language into more standardized written English, which might help them succeed in professionalized spaces. I think that’s an interesting question that potentially creates greater access and equitability.
On the flip side, however, I think it’s really important to note that much of the training data for AI is from English language writing. And so that actually cuts off access for the kinds of things that can be generated for other languages, or in particular
specific dialects or regional forms of English. And in that way, I think we want to be really cautious about the inherent linguistic bias that these tools create.
Miller: What do you think are the key factors that will determine the successful adoption of AI in
a university setting?
Myers: I think the most important thing is that users, no matter their role in the University, develop the critical literacy to understand what the limitations of these tools are, so that they can then use them in effective and productive ways. If you know it can create misinformation, then you can develop a practice of verifying information and using
AI more responsibly.
And I do think it’s important that universities don’t buy into a notion that AI is inevitable and choose uncritical adoption of these tools over the practices and assessments that help our students (and coworkers and colleagues) build critical thinking skills and foster intellectual growth. It’s been said that AI is only as good as the person using it, and I really think that’s true. More importantly, in a world where AI might begin automating anything and everything, it’s even more important that we don’t automate our own thinking and our own interior lives.
