ChatGPT in college classrooms: UT professors on uncertain future, potential positives

Since ChatGPT was introduced late last year, there have been many questions raised about how it will be used in classrooms, both as a tool teachers can use and as a means of cheating for students.

ChatGPT was introduced in November of last year by OpenAI, a research and deployment company with a mission “to ensure that artificial general intelligence benefits all of humanity.”

OpenAI has an array of features, like DALL·E 2, which produces images, and OpenAI Codex, which can turn written prompts into code. But ChatGPT is the feature making waves in the news and in classrooms. It can have a conversation, write prose, write code and generate information.

ChatGPT is not all-knowing, and there are a few things it will not do, like providing outright copyrighted works such as the text of a book.

If asked what else it cannot provide, ChatGPT will relay that it is unable to provide harmful or illegal content, medical advice or legal advice. If asked to have an opinion, ChatGPT will reiterate that as “an AI language model, I do not have personal opinions, feelings, or emotions. I am neutral and impartial.”

But the system can follow prompts, and generate specific prose that meets the distinct needs of anyone using it.

ChatGPT has become a new concern in the educational world, as it can be used to cheat.

Sean Morey, an associate professor in the English department and the director of the first-year composition program, is hesitant to immediately qualify this new technology as bad for classrooms, though.

"We don't want to be anti-AI," Morey said. "Practically every writing technology that has been invented was initially met with detractors and a huge pushback. I mean, this includes things like the pencil."

Several University of Tennessee English professors stated that they did not yet know enough about ChatGPT to feel comfortable discussing it in any certain terms.

The writing that ChatGPT generates can be used to cheat, but the writing it generates tends to be repetitive and flat, or what Morey describes as C-level work.

"It's somewhat technically correct, and it seems to do the assignment, but the facts might not be right, or it might be pulling the information from sources we wouldn't use academically,” Morey said.

Michael Knight is a creative writing professor in the English department and has so far had no issues in his classroom. He is wary of ChatGPT, but the new technology has not changed the way he conducts his classroom.

"I think sometimes we, both as humans and members of academia, go rushing toward new technologies without sort of thinking through very clearly what the ramifications of that will be," Knight said.

On Wednesday, Feb. 1, the UT English department met to discuss the implications of this new technology in classrooms. But more than warning against it, they also discussed what purpose it could serve.

"We would like to find ways to incorporate these technologies into our pedagogies,” Morey said.

As the English faculty discussed artificial intelligence overlapping their pedagogy, OpenAI announced ChatGPT Plus, which allows users to pay for a premium experience and priority access. The free version will continue to be available.

ChatGPT is already prompting responses from those trying to stop its misuse, like a student from Princeton who created an app to detect if writing was written by a bot. On Jan. 31, OpenAI announced its own AI classifier program which will detect writing done by AI. They specify in their press release that it is not “fully reliable.”

Technology will evolve, and so will classrooms, although Knight is less excited about that possibility.

"I cannot imagine a way in which ChatGPT will be good for the classroom," Knight said. "It might change the classroom, and we might have to deal with that, but I personally am incapable of imagining a way that that's going to be good.”