In Jan. 2023, Sciences Po, one of France’s top universities, sent an email addressed to all students and faculty announcing the ban on the use of ChatGPT and all other Artificial Intelligence (AI)-based tools.
Using machine learning to generate human-sounding responses, ChatGPT interacts with users conversationally and will provide a detailed response when given instructions via a prompt. It can write code and essays, solve math problems and has passed exams from U.S. law and business schools.
ChatGPT is trained on a diverse range of websites, books and text data obtained from the internet, yet there are currently no federal laws that regulate the use of personal data when training machine-learning models in the U.S.
Universities and schools throughout the globe have taken similar steps to Sciences Po since the online chatbot, created by AI developer OpenAI, was launched in Nov. 2022. The online chatbot is now banned out of plagiarism and misinformation concerns in jurisdictions throughout Australia, Bangalore’s RV University, New York City Public Schools, Seattle Public Schools and Los Angeles Unified School District.
Point Loma Nazarene University’s various departments have had discussions on ChatGPT’s effects on teaching and learning and share news updates on AI-based tools internally. However, the university has not released official school-wide guidelines on the usage of the online chatbot and other AI-based tools in classrooms.
On Thursday, Jan. 26, PLNU’s Center for Teaching and Learning (CTL) held a Zoom meeting giving a broad overview of the online chatbot. The 50 faculty in attendance were taught how to AI-proof assignments, discern an AI voice from a student’s voice and create class-wide statements about ChatGPT.
Director of the CTL, Jo Clemmons (Ph.D.), said that while PLNU is still navigating the online tool’s effect on education, the university will eventually develop new pedagogies. Clemmons also said that it is too early to know if there has been an increase in academic integrity cases where students have allegedly used AI at PLNU.
“Faculty will not just have to figure out how to detect the online chatbot, but also become creative with it,” Clemmons said.
Despite raised plagiarism concerns, Benjamin Mood (Ph.D.), assistant professor of computer science, explained how ChatGPT is not the thinking machine that people think it is; he equated it to asking a classmate for help on an assignment.
“When it comes to programming, [ChatGPT] does not always perform the task. It may produce something that looks correct but actually isn’t. This is because it is trying to resemble the answer that it is supposed to be,” Mood said.
This is partially due to the fact that the online chatbot’s knowledge cutoff date is September 2021. Mood added that Stack Overflow, one of the biggest computer science question-and-answer websites, has temporarily banned ChatGPT because it would occasionally produce incorrect answers.
Fourth-year international development major Luc Gamaunt A PLNU student admitted to being suspicious of ChatGPT and said that he would not just copy and paste answers from the online chatbot into an assignment. However, he said that it has proven to be a useful tool for understanding concepts.
Gamaunt recalled using ChatGPT when writing an essay for a class, and although the answers were not entirely correct, the online chatbot gave him ideas as to what to write about.
“It’s like having your very own spark notes/assistant tool for whatever task is at hand, but if you take something produced by the AI and mark it as your own then that’s a problem and is a form of cheating,” Gamaunt said.
Gamaunt does not think that the university should limit or restrict student access to AI-based tools, as it is the students’ fault if their learning is negatively impacted.
“It definitely can start to get tricky when students start using it to cheat or plagiarize, but just like any other method of cheating, you’re only really hurting yourself,” Gamaunt said. “If you make it all the way through college solely off the back of a chatbot, you’re probably not going to do so well in life post-grad. It’s on us as students to use the tools we have access to appropriately.”
For professor of Film Studies James Wicks (Ph.D.), much of the ChatGPT texts he has encountered have failed to approximate his students’ character, rationale and ethos.
“The type of writing I require in class, such as personal reflections on class content and class discussion papers for group discussion, would not serve students well if they used ChatGPT, so it would be rather pointless to use it. Although, I can see how ChatGPT could be a method of cheating when generating other types of writing,” Wicks said.
Clemmons added that if a student uses AI to do their work, it will shift and possibly weaken their writing and critical thinking skills.
“But we don’t really know how it will affect their skills in the long run,” Clemmons said.
To combat the potential student dependence on AI-based tools, Wicks thinks that assignments should be structured in a way where students can give personal insights, yet are required to do their own work.
“It makes me think that all students need to be writers who are as good as or better than ChatGPT. Students now more than ever need to know how to write critically, clearly and logically,” Wicks said.
According to Wicks, ChatGPT’s presence could decrease the number of assignments that might be seen as busy work.
“This changes everything forever. We academics in the workforce will be using AI for the rest of our lives,” Clemmons said.
On Thursday, Feb. 9, Clemmons will host a roundtable discussion with PLNU faculty to further discuss ChatGPT’s impact on learning.