Who’s Afraid of Woolfbot? Professor Megan Quigley and MA Student Tate Gibbons Explore the Use of A.I. Chatbots in the Classroom

Who’s Afraid of Woolfbot? Professor Megan Quigley and MA Student Tate Gibbons Explore the Use of A.I. Chatbots in the Classroom

By Nikola Stojcic MA ‘28

For many of us in the humanities, the encroachment of artificial intelligence usage into our academic lives has provoked feelings of wariness, to say the least. Indeed, what role is there, if any, for artificial intelligence in the field of English literature? For Professor Megan Quigley and 2nd year English MA student Tate Gibbons, the answer might look something like their custom chatbot, “The Lighthouse”--or, as it has come to be known, Woolfbot.

In her forthcoming article “Refined Out of Existence”: Modernist Impersonality Revisited,” Professor Megan Quigley describes how in the summer of 2024 she began studying the recently unarchived eight volumes of T.S. Eliot’s Complete Prose. To help her sift through nearly 7000 pages of prose and to better understand and evaluate the usage of A.I. for herself, Professor Quigley created a custom Generative Pre-trained Transformer (GPT) she called Scofield 1.0, named after Eliot’s childhood friend and publisher of The Waste Land. Trained on the work of T.S. Eliot scholars, gender and race theorists, as well as literary theorists more broadly, Scofield would scan Eliot’s volumes to identify writings relevant to Professor Quigley’s research interests and offer its interpretations for the selected text. Importantly, Scofield was a closed bot, meaning it couldn’t share any of the data that Professor Quigley had inputted to outside sources. The project yielded insights regarding the limitations of A.I. collaboration, as both Scofield 1.0 and its successor, Scofield 2.0, analyzed Eliot’s work with the racial and gendered biases of the scholarly works it was trained on. For Professor Quigley, the next question became how these findings might prove useful in investigating the use of A.I. in the classroom.

In a zoom call, English M.A. student Tate Gibbons told me how Professor Quigley approached him in 2025 to help develop a custom chatbot for pedagogical purposes. “She pitched it as getting our hands dirty with A.I.” Tate says. “It became a way we could deal with it and not put our heads in the sand.”

The two received funding via a VITAL (Villanova Institute for Teaching and Learning) grant to participate in a ChatGPT pilot program in 2025, which, per a report written by VITAL graduate assistant Wittney Hammeke, aimed to “explore innovative, pedagogically grounded ways to engage students through the intentional integration of ChatGPT tools into their courses.” To this end, Professor Quigley and Tate designed The Lighthouse, aka Woolfbot, for use in two of Professor Quigley’s Fall 2025 classes–one undergraduate and one graduate–both focused on Virginia Woolf’s work. Professor Quigley and Tate introduced the bot as a homework assignment for the undergraduate students and as an in-class exercise for the graduate students. The Woolfbot project was set up to be optional for both classes and included a bibliography of “Critical A.I.,” with articles explaining the potential threat of A.I. to the humanities. Students could engage with Woolfbot or write a short piece explaining their critique of engaging with A.I. in the literature classroom. The aim for both groups, according to Professor Quigley’s reflections in the VITAL report, was to see “whether students’ engagement with the bot would elicit the kind of extratextual reflections—on language, self, consciousness, etc.—that might deepen their understanding of Woolf’s fiction.” 

So what does interacting with Woolfbot look like, exactly? Its role, Tate says, is that of a theoretical “close reading partner” that will prompt you to question the meaning of a passage and to consider deeper meanings of the text. And while the bot was not in any way designed to encapsulate Virginia Woolf’s personality, that doesn’t mean it’s without a personality of its own. Indeed, the bot has an “ostensibly scholarly” persona, which is “supposed to sound like if you asked someone a question in a seminar and they gave you a three paragraph response that they had written in advance.” Additionally, although Tate and Professor Quigley made attempts to tone down the trademark “back-patting” behavior of ChatGPT, it is still prone to commend you on your expert analysis. It might also get passive-aggressive if you happen to question its authority (after testing and purposely needling the bot myself, I can confirm this).

Initial reactions from students using the bot apparently ranged from “true appreciation for extended ‘personalized’ debate to horror about talking to a bot about life and death.” But setting existential questions aside for the moment, did Woolfbot achieve its purpose on a pedagogical level? The answer depends on who you ask: “In general, for both [undergraduate and graduate] classes, there were some people who thought it was useful as a critical A.I. exercise and then fewer people who thought it was useful as a literature exercise,” Tate says. In particular, the graduate class felt the bot’s grasp of Woolf’s work was not yet solid enough to provide any literary benefits; however, many of the grad students nevertheless found the exercise to be enjoyable and a helpful illustration of the critiques and questions facing use of A.I. in literary studies.

For Tate, the opportunity to train the bot using the most advanced model of ChatGPT available at the time provided a learning experience of a different sort. “It is able to make some of these kind of fine argumentative moves that really feel like something you would see in a graduate paper, which is kind of unnerving and as a grad student, it’s not something you want to see an A.I. is capable of,” he says. “I wouldn’t have thought A.I. was capable of that if I weren’t working with these advanced models.” 

So what is the future of The Lighthouse? When I followed up with Professor Quigley over e-mail, she notes future classroom uses of the bot would continue to be optional, with those opting out required to write a critical article about their decision to do so. She says this “has proven a very useful exercise as people give voice to fears about data centers, unemployment, plagiarism, etc.” Additionally, she and Tate would revise the bot to ensure its responses are framed in terms of questions, rather than answers, to better elicit students’ own analysis.

Asked about what she hoped students took away from the assignment, Professor Quigley says, “I hope that they thought long and hard about large language models, data sets, and how their approach to reading as students might differ from the way A.I. analyzes a text.”

Professor Quigley and Tate presented their work on The Lighthouse at the VITAL 2026 Teaching and Learning Strategies (TLS) Program, as well as at a faculty retreat. Professor Quigley’s forthcoming work includes an article on speculative fiction and AI for the Cambridge Companion to Literature and A.I, and a book project “on fan cultures and reading modernism which connects to the work of Impersonality and A.I.” She is currently on sabbatical.