Muhlenberg’s reaction to ChatGPT and other artificial intelligence tools
Faculty and staff met this semester to discuss the possibilities and limitations of this rapidly advancing technology, as well as good teaching methods that can counter the perceived threats.
Posted By: Megan Kita
Tuesday, April 18, 2023 09:55
Joshua Barsczewski, assistant professor of English and director of the writing program, talks with a university writing tutor in the writing center. Photo ’23 by Joe Romano
You’ve probably seen breathtaking headlines about how advances in artificial intelligence (such as ChatGPT) make it easier for students to cheat and harder for teachers to cheat. His Joshua Barsczewski, assistant professor of English and director of his writing program, isn’t worried yet.
“People who have used it so far aren’t too worried about it because there are pretty strict limits on what they can do,” he says. “When you ask a question, it looks through the internet’s archives to 2021 to see how others have talked about the topic. Out of that comes compelling prose … with content-level expertise. Those who are familiar with it should be able to read what this produces and understand that this is a superficial analysis and is the same as what people have been saying for decades.”
Barsczewski held a workshop on tools and teaching writing for the Muhlenberg community earlier this semester. There, he and other faculty members discussed strategies for preventing this kind of plagiarism and ideas for how to use these new tools in the classroom.
Gathering of student writing tutors
For example, only certain kinds of assignment prompts might be convincingly addressed by AI. Assignments that ask students to reflect on classroom discussions, deal with ambiguous or recent texts, or revise previous assignments written in response to detailed feedback from teachers are all examples of AI’s current capabilities. Out of range.
One way to use the technology in the classroom, Barsczewski says, is to type prompts into an AI tool and discuss with students what works and what doesn’t work with the responses the AI tool generates. said. In general, AIs are not good at thinking deeply, but they are good at writing clear sentences, so the output could be used as an example of how to structure a paragraph or an essay.
The university is unsure about how faculty will use these tools in the classroom, so Gretchen Gotthard, dean of academic life, said she is holding off on developing a formal policy on their use. increase. More workshops may be held so faculty can continue to discuss these tools. The writing center lead tutor also held a workshop this semester for students who work as tutors and her writing assistants. Head tutor and writing her assistant Declan Kelsey ’24 (pictured below) said: My two majors are history and Italian studies.
“ChatGPT is still very lacking when it comes to incorporating real analytics and sources,” he says. “It can be used for short answers and discussion posts, but for long essays it’s too limited for a good read.”
In the meantime, Writing Center faculty and students will continue to monitor rapidly changing technology and how it will affect their work.
“What this technology does really well, like summarizing a piece of text, is a skill that is taught at the high school and middle school level,” says Barsczewski. “In a few years, we plan to get those students. It will eventually become our problem. We can’t bury our heads in the sand.”
