Pitt humanities professors are reckoning with a potential decrease in students’ ability to read and are adapting their curriculums to the age of AI text summarization.

College student literacy is declining, according to professors at some elite colleges, who are straying from assigning whole books. Pitt humanities professors are also witnessing this trend, correlating it with higher AI use and changing their syllabi.

Clark Chilson, an associate professor in the religious studies department, said he believes some shortcomings of the American education system, like the decrease in teaching children how to deeply read a text, may be a reason why there is a possible decrease in students’ ability to read.

“In the American school system, reading instruction typically stops in elementary school, and so students have very little knowledge of how to read in different ways,” Chilson said. “Then they go to AI, and have it explain the text in a much simpler way.”

Chilson said his concern for students’ ability to understand complex texts has increased in recent years.

“Before, when students didn’t read, I thought they were choosing not to. Now I’m actually concerned that a significant number of them cannot read,” Chilson said. “[I’m concerned] they can’t spend 40 minutes with a text to see how it [all] works together and be able to understand the inner workings of the text.” 

Chilson elaborated on his perception of students not completing class reading.

“I’ll have [students] write down discussion responses anonymously, and I’ll collect it, and 85% of students don’t look like they’ve even attempted the reading,” Chilson said.

Chilson said he started using AI in his classes in 2023 because he was willing to explore its possibilities. He also took a seminar for Pitt faculty in 2023 which focused on the integration of AI into pedagogy. The attendees created assignments with AI during this seminar, and afterwards, Chilson tried to integrate AI into his own classes, but he did not see success.

“What I thought was going to happen was that students would use AI to go deeper into the text and materials we were covering,” Chilson said. “That wasn’t happening.”

Chilson said he moved away from using AI in his classes in 2024 and has since emphasized writing assignments by hand and in class.

“I wanted to hear [the students]. I’d rather get something poorly written that was thought out than what you get from an AI generator,” Chilson said.

Liann Tsoukas, a teaching professor and director of undergraduate studies in the history department, said she doesn’t have a firm AI policy in her classroom because she believes there is no firm guidance on how AI should be used.

“I think that policies are intended to be firm, and we don’t have any firm guidance yet on this,” Tsoukas said. “I think it has to be fluid. I stay in conversation with students.”

Tsoukas said she has adjusted the length of the readings she assigns from 100 pages to about 30 and has also altered the way she teaches in response to these changes.

“I have to find different types of readings that connect [with students] and compel them to stay with it, to stay focused,” Tsoukas said. “We’ve gone from being focused and engaged to kind of erratic with lots of interruptions and declining focus.” 

Tsoukas said the overall academic expectations some professors have for students has lowered in recent years. 

“We [have been providing] them the way to think about things rather than saying, ‘How do you want to think about this?’” Tsoukas said.

Tsoukas introduced a new policy in her U.S. history class last semester to compromise with students. She assigned a memoir on a civil rights activist but noticed many students did not read the book. As a result, Tsoukas developed a new strategy.

“Rather than have students read the book and then write an essay, we told them there was an alternate option,” Tsoukas said. “We broke up the book into chunks, and every time they came into lecture there was a six question quiz about that brief chunk.”

Tsoukas said technology’s potentially negative influence on students’ ability to learn history has made teaching the subject more difficult.

“Understanding the complexities and the complicated truths of [history] is through patient engagement with sources,” Tsoukas said. 

Jennifer Waldron, an associate professor of English, said she believes Shakespeare is harder for current students to understand because some high schools have stopped teaching whole plays and started prefacing them with their respective film adaptations. However, this does not mean her students have stopped their interest in learning Shakespeare.

“This means that when they come to read Shakespeare at the college level, it’s more of a challenge,” Waldron said. “They’re coming [to class] curious, open-minded and interested, but they have not had as much practice tackling complex texts.”

Waldron said she believes that with the increasing use of ChatGPT, students are getting more used to reading literature summaries, rather than doing close reading, which she said English professors prioritize. 

“With modern digital communication, and with ChatGPT, people are reading more for information,” Waldron said. “They’re trying to get the core idea out of something quickly, as opposed to valuing the process of immersing themselves in the text.”

Waldron said a key element of Shakespearean literature is his invention of words, which should be analyzed through close reading rather than skimming. Another feature of Shakespeare is his use of metaphors, which Waldron said ChatGPT cannot convey deep meaning from.

“We want to understand how [Shakespeare] invented those words in the cultural context, not just sort of translate them into a modern word that might mean the same thing,” Waldron said. 

“He makes meaning through these larger structures of how the different details go together, not just what metaphors mean literally.”