USC faculty and students explore AI’s evolving role in education during symposium

The study session devolved into an argument.

One student said he uses ChatGPT regularly for assignments and turned to it for a report on the novel Frankenstein and how it compares to director Guillermo del Toro’s spin on the classic tale in his recent film.

“AI is just a tool,” he explained as the trio snacked on chips, an apple and energy drinks. “I asked questions, and I took the answers deeper.”

One of his study partners said she wrote her paper herself without the aid of any AI chatbots and therefore had a deeper understanding of the material.

The third student said he could relate to both.

Convinced, however, that many university students aren’t using AI ethically, the woman left the study group, saying her human input apparently would be of no use.

She viewed using a chatbot to help write a paper as an exercise in intellectual laziness as opposed to using one’s own brain.

Brainstorming with others and human connection, in her view, is critical to strengthening one’s intellect.

Although the study session was a skit featuring students from the USC School of Dramatic Arts, the scene rang true for attendees of a recent presentation, “Learning, Teaching, and AI: A Community Conversation on Ethics and Higher Education in the Age of AI.”

Spirited discussion on AI

The March 31 forum in the Dr. Allen and Charlotte Ginsburg Human-Centered Computation Hall Auditorium featured a spirited discussion by panelists and audience members on the ubiquity of AI and what it means for higher education.

“Suddenly, [AI is] everywhere, seemingly infiltrating every nook, cranny and corner of academia and met with a combination of suspicion, bemusement and uncertainty,” said moderator Andrew McConnell Stott, vice provost for academic programs and dean of the graduate school.

Stott posed several provocative questions at the forum, which brought together two important USC initiatives: the President’s AI Strategy Committee, chaired by Geoffrey Garrett, dean of the USC Marshall School of Business, and the Open Dialogue Project, managed by Neeraj Sood, a professor at the USC Price School of Public Policy.

Both Garrett and Sood attended the discussion.

“Will [AI] break our educational systems by revealing their overreliance on transactional assignments and memorization tests that can be aced by [a chatbot] in seconds, or will human-mentored algorithms lead us to a new age of discovery and productivity?” Stott asked.

He noted that Sam Altman, the CEO of OpenAI, predicted that in 18 years, colleges will become obsolete.

“If AI does break our current model,” Stott asked, “what will this alternative look like?”

Student and faculty panel

Panelists at the 90-minute community forum were Mark Redekopp, professor of electrical and computer engineering and computer science at the USC Viterbi School of Engineering; Sarah Mesle, a teacher in the writing program at the USC Dornsife College of Letters, Arts and Sciences; and Ethan Osiegbu, a sophomore studying biology and public administration and a board member of the Open Dialogue Project.

The USC Open Dialogue Project, initiated last October by USC President Beong-Soo Kim, aims to foster a campus culture in which Trojans embrace and embody the principles of academic freedom, free expression and open discourse — essential values that nurture intellectual courage, open-mindedness, thoughtful engagement, curiosity and the pursuit of knowledge.

The verdict: It’s complicated

So, is AI good, bad or complicated?

The panelists agreed it’s complicated.

“I’m anticipating a lot of opportunities with AI but a lot of pitfalls,” Redekopp said. “It does such a good job that a lot of students are asking themselves, ‘What is the value of the classroom? What does the classroom add [to my education]?’”

“I think that’s the challenge I feel as an educator,” Redekopp continued. “How do I adapt and bring something of value that augments what students can get from other sources?”

Redekopp, who has taught for 20 years, said he’s more excited than he’s been in the past to imagine what the classroom can be as AI becomes more entrenched in higher education.

Still, he said: “I want students to go through learning experiences, and AI sometimes allows them to bypass the intellectual steps they need.”

Redekopp noted that since AI took hold, the number of interactions he’s had with students during office hours has plummeted 80% to 90%.

Stott agreed.

“My office hours are some of the loneliest times of my life,” he said. “It’s getting bleak out there.”

Banning AI in the classroom?

Mesle said she’s also excited that the classroom experience is changing.

“The sense that a revolution is happening is interesting to me,” she said.

But, she declared: “ChatGPT can’t help my classes. Writing is hard, and it takes tenacity to stay with a difficult challenge. How do you become a person who does the hard thing? We need to hold each other accountable to that kind of tenacity.”

Mesle bans digital technology in her classroom.

“We talk to each other and persist with the awkwardness of human interaction,” she said.

Osiegbu said the real problem might not be AI, but the university.

“You put your best foot forward on an assignment [not using AI] and some students who use AI get a better grade than you, which is pretty annoying,” he said.

“If there’s a means to take a shortcut,” Osiegbu continued, “students will take it. I think this is the fault of the education system putting an emphasis on passing as opposed to innovative and creative learning. People have gotten so bogged down by getting a good grade to get to the next level.”

“I know I need an ideal GPA, but maybe this is more about the educational system as opposed to AI … maybe it’s the structure of education that needs to change.”

The personal trainer analogy

Redekopp bemoaned students who turn to AI as a crutch when the going gets tough.

“Some of them get to a point where they come to a potential solution and then it doesn’t work,” he explained. “But that’s the point where they are ready to figure things out and learn. The temptation then is to use AI.”

“An analogy I like to use with my students is that of a personal trainer. As soon as something got difficult, if the personal trainer jumped in and lifted the weights for you, you would be appalled. And you wouldn’t want to pay your trainer for that,” Redekopp said.

“If you use AI in the same way,” he continued, “you’re never going to build the mental muscle you need to solve harder problems in the future.”

‘We need to be goofy’

Stott wondered whether AI will one day be viewed like electricity, which didn’t happen overnight but eventually became a “normal” technology.

“If AI is going to be normalized, it needs to be intensely regulated,” Mesle said. “And the role of a university isn’t to race to adopt the latest technology — that’s not leadership, that’s following.”

Mesle said the reliance on chatbots and the generic-sounding prose they generate poses a risk to students’ voices coming out in their writing.

“I’m extremely worried about that,” she said. “I want all of us to feel we have agency and are brave. We have a lot of power and should claim that. We need to be goofy.”