Nick Zufelt hosts the second installment of the series regarding the intersection between AI and education.
Exploring how ChatGPT can be used to enhance learning experiences, Nick Zufelt, Instructor in Mathematics, Statistics, and Computer Science, led a “Lunch and Discussion” event on November 6 in the Tang Gallery. In addition to a presentation, the event featured an activity where participants engaged with artificial intelligence (AI) using different prompts. This event is the second installment of a series hosted by the Tang Institute, dedicated to examining the intersection of AI and education.
Inspired by the recent research paper “Assigning AI: Seven Approaches for Students, with Prompts” by Ethan Mollick and Lilach Mollick, Zufelt expressed that conversations around ChatGPT should be expanded to explore how AI can support rather than supplant a student’s thinking. He hoped to provide a space for students and faculty to experiment with the roles that AI can play in learning.
“It is really important to try and expand the dialogue on how we’re talking about generative AI on campus. There’s a lot of conversation right now about how generative AI might replace someone’s thinking, and then that’s when we start getting into academic integrity concerns,” said Zufelt. “I think those are extremely valid concerns.”
Zufelt continued, “But I’m actually really interested in exploring a totally different avenue about how AI can expand or augment one’s thinking. Rather than the [user] giving AI a task so that they don’t have to think, [they could see AI] as more of a partner in the thinking process.”
Zufelt allows students to use ChatGPT in his own classroom but requires students to fully understand the code they submit. He compared the interaction between a student and an AI to the interaction between a student and a tutor.
“I do ask students to code alongside a generative AI in a computer science classroom, but I tell them, and I regularly check in with them, you have to understand everything that you turn in… Rather than never using ChatGPT or having it replace their thinking, [students can] think about it as… [talking] to somebody at a computer science study center. A tutor would give them the information but also explain why it’s working, so that hopefully the next time it comes up, they can figure it out on their own,” said Zufelt.
Emma Capaldi ’25, an attendee of the event and a student in Zufelt’s Computer Science 561 Machine Learning course, emphasized the importance of discussion as AI usage becomes increasingly normalized. She shared how Zufelt’s discussion broadened her perspective.
“The different scenarios illustrated the different ways you could use ChatGPT in a way that wasn’t, ‘Do my homework for me,’” said Capaldi. “People will use ChatGPT whether we talk about it or not in class, so I think it’s important to lay out those guidelines and give students constructive ways to use ChatGPT. Don’t tell them, ‘You cannot use any AI ever,’ tell them you can use it for these useful things that are not going to be detrimental to your learning.”
Howie Kalter, Instructor in Mathematics, Statistics, and Computer Science, appreciated the opportunity to learn about using and applying ChatGPT to the classroom. In particular, Kalter described how providing more detailed and systematic prompts could elevate the quality of user interaction.
“The main takeaway for me was how important the initial prompt is when using ChatGPT. Dr. Zufelt’s examples included paragraphs of prompting. When I used ChatGPT in the past I used only a few sentences. It was amazing that giving more detailed information can result in incredibly powerful interactions with ChatGPT. One of Dr. Zufelt’s examples created a very helpful interaction with a ‘tutor’ who helped me learn about rocket science!” wrote Kalter in an email to The Phillipian.
As AI continues to develop rapidly, Zufelt noted the concern that generative AI will replace creative processes such as writing. He commented on the importance of how humans should continue to write instead of relying solely on AI.
“Writing is something that is deeply human, and it’s not going away. One of things that is really interesting about generative AI is it’s helping us to study that whole idea. Why is it that people are so upset about that? I think it’s because writing is a fundamental human craft, and the threat of losing writing as a human endeavor is really scary. It is conceivable that, depending on how the future shakes out, writing begins to take more of a backseat, and I think that’s really dangerous,” said Zufelt.
