Teachers respond to ChatGPT

Tate Sheehy and Saisha Kumar

The History Department released an official statement prohibiting the use of artificial intelligence (AI) on assignments in response to the emergence of AI program ChatGPT on Jan. 15. As of Feb. 4, the English, Math, Science and Language Departments have yet to publicize formal language regarding the use of AI on assignments.

ChatGPT is an AI chatbot created by Open AI in Nov. 2022 It is capable of formulating essays, writing code and answering questions when prompted to do so. ChatGPT presents a threat to conventional learning because it is able to create nearly instantaneous responses to complex questions, interfering with students’ learning, according to AP News. It may also provide users with inaccurate information, according to OpenAI.

The History Department’s statement reads, “The History/Social Studies Department does not support the use of ChatGPT and other AI platforms for any purpose. The use of these platforms is considered unauthorized aid under the Honor Code; thus, any preparation for the course and any submitted work for the course that are generated through or are informed by ChatGPT or other AI platforms are in violation of and subject to consequences under the Honor Code.”

President Rick Commons said although he believes restricting the use of ChatGPT on assignments is currently possible, it will ultimately become a part of the curriculum.

“I think that we are going to have to figure out how to use the ChatGPT and other artificial intelligence in the process of teaching,” Commons said. “I don’t think we’re going to be able to long-term avoid or ban it. I don’t blame the History Department for essentially banning it for now. I don’t think they can ban it for all time. I think they’ve got to ban it for now while they figure out how to make sure that when we use it, we’re using it in a way that is effective and equitable.”

History Teacher Katherine Holmes-Chuba said using ChatGPT is cheating because the program, not the student, generates ideas.

“For people in the humanities, it ultimately is plagiarism because even though an AI is doing it, it’s not your ideas, and by going to ChatGPT, you’re not thinking about the material, and you’re just taking the easy way out,” Holmes-Chuba said. “It worries me a little because I always think that one of our strengths [as teachers] is teaching students how to think for themselves.”

Holmes-Chuba said ChatGPT detracts from a student’s educational experience.

“I understand that it can be super helpful in certain businesses and areas, but I think for schools, it’s like a big low-hanging fruit, and we are in the Garden of Eden right now,” Holmes-Chuba said. “For a student, I think it’d be very hard not to be reaching for that apple.”

Director of Instructional Technology Jeff Snapp hosted an informational session for faculty on Zoom to discuss the threats presented by new technology like ChatGPT on Jan. 18.

Snapp said ChatGPT could disturb a student’s ability to learn and develop important skills.

“Every teacher wants all of their students to articulate their thoughts, weigh evidence and refine critical thinking skills, so they develop assignments that give students the opportunity to practice becoming better communicators,” Snapp said. “The emergence of and ease of access to language-based AI threatens to provide students with a crutch that would rob students of the opportunity to practice these skills. When students make choices to go outside the parameters established for assignments and use unauthorized aid, the consequences extend beyond just the grade they receive.”

Snapp said although some students and teachers may struggle to grapple with the new technology, they will eventually realize its importance.

“New technologies have emerged throughout history and were not initially welcomed,” Snapp said. “Today, we take disposable pencils for granted, but when they emerged, people who valued the skill and beauty of writing with a quill pen expressed concern that valuable skills and jobs would be replaced. While [AI] will be disruptive and uncomfortable for some people, new opportunities and capabilities will emerge. One day, we’ll look back and laugh in wonder at how we were able to survive without our AI tools.”

Head of Upper School Beth Slattery said discussions of reducing take-home assignments prior to the introduction of AI.

“There were a lot of discussions prior to ChatGPT about take-home assignments because some students were already submitting over-tutored assignments, which is both unethical and inequitable since not everyone has access to tutoring,” Slattery said. “It really doesn’t serve anyone in the long run to cheat on take home assignments, but because we know that is happening, we may reduce those kinds of assignments.”

Science Department Head Melody Lee said the Science Department is deliberating approaches to address the rise of ChatGPT.

“I’m not so worried about students taking advantage of it, at least in my science class, just because most of the stuff is based off of data,” Lee said. “It just leads to more conversation because this is kind of like having a tutor, in a sense.We were talking about using a lockdown web browser so students couldn’t use it for in-class writing or giving assessments where students have to write in class.”

Mathematics Teacher and Robotics Coach Andrew Theiss said he thinks AI has the potential to be helpful when used in the right circumstances.

“ The last thing we should do is consider it a threat to education,” Theiss said. “We have to view it as a tool that can be used to support student learning and improve their skills. As a department, it really depends on the specific use cases, so I think that as a tool, it can help, but I think that if we can’t establish clear guidelines, it could be misused and affect student learning.”

Theiss said finding a balance between student and AI-generated work is essential to the learning experience.

“[Guidelines] would include requiring students to cite the source of any generated text by the tool and maybe setting up a word limit for the amount of text that can be generated by ChatGPT,” Theiss said. “I think those types of limits, like limiting how much you can use or leverage ChatGPT, would balance the amount we use it with trusting ourselves.”