With its potential for revolutionizing the way students engage with course materials, ChatGPT is at the center of ongoing discussions and debates among universities as they weigh the pros and cons of incorporating this advanced technology into their curriculums.
That was written by ChatGPT, an “advanced language model tool” created by software firm OpenAI.
ChatGPT — a chatbot capable of providing information on command, generating content and explaining concepts — is turning heads at higher education institutions across the country. Weeks after its November debut, a Furman University student confessed to enlisting the chatbot to complete a philosophy paper, according to Insider. Since then, academics have proven it capable of passing the U.S. Medical Licensing Exam, a Wharton Business school exam and four law exams.
Several K-12 school districts and universities have even banned the ChatGPT outright, citing fears over “plagiarism and misinformation,” according to Insider.
But UB is among a shortlist of universities exploring a future with ChatGPT in its lecture halls.
The university says its Office of Academic Integrity and Office of Curriculum, Assessment and Teaching Transformation (CATT) are collaborating to help faculty “incorporate new tools as part of the learning process” without backing down on academic integrity.
“UB supports innovative teaching, including the use of emerging technologies such as ChatGPT,” the university said in a statement. “Student use of these technologies, however, should never violate existing academic integrity policies.”
An “Academic Integrity at UB” course that was piloted this past academic year will also be made mandatory for all new students in the fall.
“Students who use ChatGPT outside of the bounds set by their instructor will be charged with academic dishonesty,” Kelly Ahuna, director of the Office of Academic Integrity, said. “We know students are likely intrigued about ChatGPT and how it might assist them. Different courses may be more inclined to use AI technology as a tool, so it will be important for instructors to set parameters around its use.”
Several students and faculty members spoke with The Spectrum about their attitudes toward ChatGPT and classroom experiences one week into the spring semester.
Zach Leng, a freshman electrical engineering major, says he got the lowdown on ChatGPT in Introduction to Programming with professor Nicholas Mastronarde by the end of the syllabus week.
“[Dr. Mastronarde] concluded that it [ChatGPT] certainly can be used to write a program, but just because the program can run without errors doesn’t mean it is actually doing what you wanted it to do,” Leng said. Mastronarde warned students against using ChatGPT but admitted that detecting AI code compared to human code would be difficult “since they can look completely identical.”
Jinhui Xu, chair of the Department of Computer Science and Engineering, says that the lack of concrete detection methods has prompted his department to increase the weight of exams at the expense of homework and projects, to compensate for the use of ChatGPT.
“I want to feel like this will help students to learn – of course, maybe they can use ChatGPT as a starting point. But if it teaches students the wrong information, that is even worse than not learning,” Xu said. “But like any other new technology — because we’re computer scientists — we are also excited about this new development because it provides a lot of other opportunities. All breakthroughs in technology provide us with opportunities to improve and make people’s lives easier.”
Leng supports banning the tool in most classes and would like to see the university adopt AI-checking tools similar to GPTZero (the brainchild of Princeton senior Edward Tian) to counter ChatGPT.
“I am sure that certain classes could absolutely utilize ChatGPT as a tool,” Leng said. “However… classes such as ENG 105 or CSE 115 are teaching students how to write, whether that be writing code or writing essays. ChatGPT would completely circumvent these course goals.”
Milind Kumar, a freshman computer science major, is taking an introductory English course. While ChatGPT wasn’t mentioned by name, he says his professor switched to paper to make it more difficult for students to cheat using online tools.
“I personally think that the switch from digital to paper was necessary, to be honest,” Kumar said. “The course challenges someone on the basis of their writing and grammar skills, and cheating your way out using ChatGPT would be unfair for students who genuinely work hard to get a good grade.”
Still, Kumar sees ChatGPT as another adaptable tool for learners to be able to help themselves in the bigger picture — given the right motivations.
“We had Google before, now there’s ChatGPT,” Kumar said. “Personally, I believe that seeking help is OK if you understand the concept by the end.”
Sinai Thomas, a senior chemistry major and former Spectrum staff writer, has other opinions.
“From an academic perspective, ChatGPT should not be allowed,” Thomas said. “Letting a machine do the learning for you? I feel like it reflects our shorter attention spans as our generations keep going.”
Thomas says taking the time and effort to learn is more important than any outcome. He says he struggled with writing in high school but was able to improve aspects of his writing with years of English courses. He even became an English minor.
“You need to have those fundamental skills in order to succeed in the future, especially if you’re planning on going to grad school,” Thomas said. “So to me, to basically choose the easy way out doesn’t make any sense.”
Thomas isn’t alone. Zachary Stapleton, a senior education major, recounted similar discussions in a “digital tech for teaching” course he had taken.
“We were talking about the risks it would pose for the classroom when we become teachers,” Stapleton said. “It was super interesting because it’s a real issue, and we had to think of ways to combat against our own students using the program to finish their assignments.”
In one class, Stapleton was told to test out ChatGPT to get a feel for the ease of access to “quick, easy assignments” and try to find commonalities between AI-generated essays or “tell-tale” flags.
“I can see the appeal,” he said. The senior noted that the tool was error-prone and seldom cited sources correctly.
This is down to ChatGPT being a “pre-trained” chatbot, according to Xu.
“Whatever is generated by ChatGPT is based on similarity to the training data, so it does not really guarantee correctness,” Xu said. “That is not something it cares about.”
“I find ChatGPT to be both a blessing and a curse,” Stapleton said. “It’s incredibly scary to think of students just telling the bot to write an essay and then submitting it. It makes our jobs as teachers hard, to make sure that it’s not AI-generated.”
But Stapleton does see silver linings. ChatGPT can help students outline essays and generate topic ideas, and is forcing teachers to be more creative with lesson plans.“It sort of forces educators to consider more project-based learning,” Stapleton said. “As we all know, we should constantly be looking for new ways to educate, and while [cheating through AI essays] is a bad reason, it is a way to start the process.”
Carol Van Zile-Tamsen, associate vice provost and director of the CATT office, says that the university is in the process of sharing both cheating prevention guidance and ideas on implementation with instructors.
“While there is a lot of anxiety about cheating, instructors now must consider how AI tools can actually promote writing instruction just as the arrival of the calculator changed how math is taught,” Van Zile-Tamsen said.
Ahuna says that instructors will be able to tell if a student’s work is “suddenly different” and may start “asking for drafts, oral components to assignments, or in-class writing to accompany assignments.”
Meanwhile, in Davis Hall, Xu teeters between skepticism and reverence.
“I have mixed feelings,” Xu said. “It is certainly a challenge, but there’s also an opportunity for us to improve our education and our lives.”
Kyle Nguyen is the senior features editor and can be reached at email@example.com
Kyle Nguyen is a senior news/features editor at The Spectrum.