Graphic Courtesy of Evelyn Shaw
St. Ambrose computer science major George Coin continues his four-part series, “AI at Ambrose,” on how AI is challenging students and faculty. In the second segment, he explores how professors are adapting their teaching and assignments to AI and how students experience those changes.
Dr. Tanya Randal’s college students have something in common with her third-grade son: handwritten workbooks.
The philosophy and critical thinking professor says that she borrows the format to keep students working through readings themselves as AI makes summaries readily available. Her students annotate passages, read aloud, and explain ideas in their own words. But she says the approach brings a frustrating tradeoff: time spent checking completed workbooks can replace the conversations she wants the reading to spark.
For faculty at SAU, responding to AI involves choices about how class time gets spent and what students should do independently. Students feel those changes, too, and some question whether efforts to prevent cheating make it harder to learn.
Bringing the Work Back to Class
Randal, co-leader of SAU’s AI Institute, says AI arrives on the heels of pandemic-era online learning, leaving faculty to untangle two disruptions at once: Students need more help reading college-level material, she says, just as chatbots make it easier to skip that reading altogether.
“I can only blame so much of it on AI.”
Her response brings reading into the classroom, where students read aloud and paraphrase together. That gives her a chance to work through difficult passages with them, but it also changes what she can teach. Time spent helping students understand a text leaves less room for developing their writing, a responsibility she says she believes should extend across disciplines, beyond English classes.
“I can’t teach you to read and teach you to write and teach you philosophy,” Randal says. “All at the same time.”
The adjustment carries a cost: She says she spends more time guiding students through the reading that once prepared them for class, leaving less time for the conversations she wants that reading to start.
Randal says she has also reduced her own use of AI for course content. She says fixing unsuitable examples and images in AI-generated presentations can consume the time the tool is supposed to be saving.
She draws a firm line at using AI to assign grades. A chatbot can give students feedback on a draft, she says, but handing over grading distances professors from their students’ work.
For students, she explains acceptable use within individual assignments. AI may help locate sources, which students must still read and evaluate. Using it to write a journal entry, however, bypasses the thinking the assignment is meant to develop. “I think that especially when we’re in a learning situation, we need to emphasize process over product.”
That emphasis on the work behind an answer is familiar to junior psychology major Marianne Hammar, who says her professors devote more class time to writing and require more assignments to be completed by hand.
“It definitely means assignments take a lot longer,” she says, though she does not find them more difficult.
Hammar says AI policies vary across subjects. Most of her psychology courses prohibit AI, while others take a more relaxed approach. She says she wants professors to teach students how to use AI responsibly while setting clear limits and worries that concerns about cheating can cast doubt on work students complete themselves.
“It’s only going to become more prevalent in society, and so I think it’s important to be taught how to use it as a tool instead of just abusing it.”
A Conversation With the Past
SAU history professor Dr. Ethan Gannaway says he also has had to adapt his classes to focus more heavily on in-class assignments. He says he no longer assigns take-home essays, which traditionally would give students more time to comprehend the material and develop their thoughts.
“It makes me really, really sad about it because I like to give them more time to process thoughts,” he says.
But AI also opens up experiences Gannaway could not otherwise offer his students, including the chance to question figures from the ancient world.
One of Gannaway’s favorite uses of AI lets students question Julius Caesar. He supplies ancient biographies to ground the chatbot in primary sources, then gives students room to pursue their own questions about the Roman leader.
“They can ask questions following their own line of inquiry,” he says. “And the students who especially engage it are doing that critical thinking of responding in real time.”
For Gannaway, the exercise depends on what students do after the chatbot answers. Can they consider its meaning, challenge it or ask a question that takes the conversation further? He says those are skills students need beyond the classroom, too.
“We have to be so deliberate about how it’s used,” he says. “Otherwise, they do let it do all the critical thinking for them too.”
When Gannaway suspects AI-written work, he says that he asks the student to explain their ideas and word choices, using the conversation as an opportunity to discuss appropriate use and what students risk losing when they rely on a chatbot.
“You’re losing out on your voice and your thoughts and your curiosity,” he says.
He would rather students struggle to express an idea than hand over language they cannot explain.
“I try to keep telling them really what I want is to hear your voice.”
That question of voice turned into a lesson of its own in an assignment senior English and public relations double major Sami Kelting describes. She says her class compared poems by writers such as Shakespeare and Edgar Allen Poe with AI-generated imitations, examining how a chatbot reproduces a poet’s style and where the writing differs in creativity and expression. For Kelting, the comparison offers a way to explore what she calls “the soul of writing.”
Kelting says professors also help by showing students how to construct prompts and refine questions for research ideas. She says these lessons give her a clearer sense of where AI can support her own work.
“I think that can really help me, like understand how I can use AI alongside learning,” she says.
Checking the Chemistry
Chemistry professor Dr. Andy Axup draws a different boundary. He says that homework is for practice and students are free to consult classmates, tutors, or AI. In-class exams show what they can do independently.
“We want you to succeed on that; we want you to learn from that,” he says of homework. “And if you use AI that’s fine.”
Axup says the chemistry department changed lab assignments after suspecting students were submitting AI-assisted reports. Students now collect and analyze data in class and write shorter explanations of their results before turning them in.
Axup also tests AI’s answers himself, and he says its mistakes in organic chemistry should give students a reason to be careful.
“ChatGPT is not an organic chemist and makes a lot of mistakes.”
In one lesson, he describes presenting an incorrect AI response without initially identifying its source. He then worked through it with his students by drawing out the molecule from the name AI provided and then comparing it with the actual structure. They did not match.
Axup jokes that he is bilingual in English and chemistry. For students learning that second language, an AI response can provide another example of how to explain ideas using chemistry’s vocabulary, much as a textbook, professor, or tutor would.
The same response can help or hinder learning, he says, depending on what students do with it. Reading carefully and working to understand the explanation can help them develop their own. Simply turning it in defeats the purpose.
Concern that AI lets students skip that practice also shapes what professors assign and grade. Sophomore exercise science major Evan Bratten says one of his science courses bases grades on tests because the instructor worries students will use ChatGPT on assignments. For Bratten, that approach leaves a gap in the practice he relies on to learn. “I like having like a practice worksheet or something,” Bratten says.
He says he values those assignments both for points and as study material. Without them, he says, “I’m kind of just left on my own.”
Bratten says he worries that efforts to prevent AI misuse go too far, taking away the practice and guidance students need to learn the material.
Making the Thinking Visible
Fr. Robert “Bud” Grant says students can use AI in his courses, provided they acknowledge it and trace the information to its original sources. Fr. Bud says the difficulty often begins when students hide that use.
He says students can feel policed while he worries they are trying to get away with something. He says telltale signs such as formulaic phrasing and suspicious language that sounds unlike the student can prompt a conversation, but spotting those patterns is easier than proving AI was used.
“It’s harder to prove it,” he says.
He describes a meeting with a student who brought their computer to demonstrate how they used AI. He says the approach itself was acceptable, but the lack of transparency and disclosure was not.
“All you needed to do was say up front that you were doing that and you wouldn’t be in my office right now,” Fr. Bud says.
He recalls another student who spoke English as a second language using AI because they lacked confidence writing in English. Fr. Bud invited them to submit their work in Spanish, which he can largely read through his knowledge of Italian. He says this experience shows both the value of language assistance and the risk that AI’s rewriting can obscure the student’s own voice.
Fr. Bud does not use AI to prepare lessons or quizzes, but experiments with theological questions in his own study. Unsatisfying answers push him to refine the question and examine his reasoning.
“When I start asking the right questions, then I’m actually learning myself.”
That engagement is what he wants to see in students’ work, too.
“We’re not interested in your answer, we’re interested in how you got to that answer,” Fr. Bud says.
That expectation of visible thinking extends to faculty, too. Senior art history major Grace Rood says she wants professors to disclose AI use and remain willing to revise what it produces.
“It then shows that they know that it could be flawed, and that they can change it.”
Rood says using AI to grade students or generate an entire course syllabus concerns her.
Across these classrooms, professors want to see students work through ideas, and students want the guidance that helps them do it. Both want the other to stay engaged in the learning process, even when AI can do the work.
Randal’s paper workbooks capture that tension. She says they help her see that students are doing the work, but she still wants more time for the conversation that follows. As faculty keep adjusting to AI, that exchange remains what both sides are asking for: a student with something to say, and a professor ready to work through it with them.
In the November Buzz issue, George will explore how people use AI in their personal lives, from medical questions and relationship advice to companionship, and the privacy concerns that come with sharing personal information. Read the first installment, “AI at Ambrose: Who’s Doing the Thinking?”, to learn how students use AI in their studies and where they draw the line on cheating.
