Office Hours: Faculty Discuss the Impact of Generative AI on Higher Education

Office Hours: Faculty Perspectives on Generative AI in Higher Education

As part of our regular Opinions series, “Office Hours,” we strive to showcase a variety of faculty voices reflecting on the implications of recent technological advancements in higher education. Our latest edition focuses on generative artificial intelligence (AI) as it relates to the liberal arts and the educational landscape at Swarthmore College. We gathered insights from several professors who shared their thoughts on this evolving topic.

The Reality of Generative AI

Syon Bhanot, an Associate Professor of Economics, asserts that generative AI is an inescapable reality in both student and faculty environments. Instead of resisting its integration, he advocates for adapting educational practices to enhance learning. Bhanot suggests "AI-proofing" assessments to ensure they remain rigorous while also recognizing that AI can serve as a valuable tool for specific tasks, such as coding. He envisions a college-wide strategy that incorporates AI in a thoughtful manner, aiming not only to improve efficiency but also to alleviate the burdens of overworked faculty.

The Challenges in Translation and Interpretation

Sibelan Forrester, the Sarah W. Lippincott Professor of Modern and Classical Languages, brings a nuanced perspective to the discussion, particularly concerning translation and interpretation. While acknowledging that basic AI has been around in the form of dictionaries and phrasebooks, she highlights the limitations of current generative AI in grasping the subtleties of nuanced language. Forrester notes the challenges of translating less commonly spoken languages, cautioning against over-reliance on AI for tasks that require a deep understanding of cultural context and metaphor.

A Skeptical View of AI’s Role in Scholarship

Emily Gasser ’07, Associate Professor of Linguistics, takes a critical stance on generative AI, labeling it as merely “synthetic text extruding machines.” She highlights that while LLMs (large language models) can generate text, their lack of understanding and insight renders them ineffective for rigorous scholarship. Gasser emphasizes the educational value of critical thinking and creativity, arguing that outsourcing these intellectual exercises to AI diminishes the learning experience. She warns that reliance on AI tools may foster a dangerous complacency among students, undermining the purpose of their education.

Cognitive Effects of AI

Sam Handlin ’00, Associate Professor of Political Science, acknowledges the potential benefits of engaging with AI but remains deeply concerned about its cognitive implications. He points to emerging research suggesting that using AI for complex tasks can impair cognitive function. Handlin argues for a college-wide policy to delineate the appropriate use of AI, emphasizing that students must exercise their cognitive abilities to prepare for the demanding academic environment at Swarthmore and beyond.

Concerns Over Learning Deterioration

Emad Masroor, Visiting Assistant Professor of Engineering, raises alarm about the detrimental effects of generative AI on student learning. He views these tools as shortcuts that lead to a superficial understanding of knowledge, neglecting the value of deep cognitive engagement. Masroor echoes the sentiment that true learning arises from grappling with challenging tasks and warns that students may enter the job market lacking genuine skills if they rely on generative AI to circumvent difficult work.

The Dual Nature of AI’s Utility

Donna Jo Napoli, Professor of Linguistics and Social Justice, offers a more tempered view, recognizing AI’s usefulness while stressing the importance of critical engagement. She argues that while AI can assist with mundane tasks, it must not replace the deeper intellectual engagement that Swarthmore students are meant to pursue. Napoli encourages students to explore their interests without shortcuts, emphasizing that true comprehension comes from the rigorous process of inquiry.

Navigating AI in Assessments

Federica Zoe Ricci, Assistant Professor of Statistics, embraces the potential of generative AI tools while recognizing their pitfalls. She highlights the need for students to develop the ability to interact with AI without compromising their educational experiences. Ricci calls for an approach that incorporates AI as an adjunct to learning rather than a substitute, backing her stance with a focus on in-class assessments that demand independent thought and preparation.

The Role of Writing in Education

Warren Snead, Assistant Professor of Political Science, voices concern over AI’s impact on essential writing skills. He argues that the very act of writing should be a difficult process that encourages critical thinking. According to Snead, reliance on AI for writing tasks detracts from students’ ability to articulate ideas and engage thoughtfully with their subject matter, which is a crucial component of their education.

---

This compilation of faculty perspectives showcases a rich tapestry of thought on the implications of generative AI in academia. Each professor presents unique insights into how these advanced tools can both enhance and undermine the educational experience, pushing us to consider how best to navigate this complex landscape.