John Warner argues that to fully appreciate the importance of the writing process, we must move away from treating education as a mere transaction.
“We have to renew our relationship to the experience of work, to be more in touch with the process, rather than just generating the product,” Warner said. “Education is more than a transaction. If we can decide what’s good or not good with a human sense, we don’t have to worry about ChatGPT or large language models taking over.”
On Oct. 8, Warner, a writer and former professor at College of Charleston and Virginia Tech, among other institutions, delivered the second lecture in the annual Nelson Distinguished Speaker Series at Harvey Mudd College. Titled “Resist, Renew, Explore: Human Values in the Age of AI,” the talk explored the importance of preserving the humanity of the writing process amidst the rise of generative AI. Warner, author of the upcoming book “More than Words: How to Think about Writing in the Age of AI,” has written extensively about writing pedagogy and higher education.
Kyle Thompson, the director of learning programs at Harvey Mudd College and director of the Nelson Speaker Series, chose Warner as the speaker in order to present multiple perspectives on this year’s topic, AI and learning. He highlighted Warner’s focus on day-to-day teaching and writing pedagogy, contrasting it with the “business standpoint” of previous speaker Sal Khan.
“Warner himself has been critical of how Khan [frames] technology as potentially saving education,” Thompson said.
Warner emphasized that writing articulates the process of thinking.
“When we write, we are trying to express an idea,” Warner said. “While you are writing an idea, it is taking a different shape. We have to see writing as an iterative sequence and recursive [to] earlier parts of the process.”
He dispelled the common belief that writing is linear, arguing that this fundamental misconception reveals why AI cannot truly replicate the process.
“The linearity of writing is illusion after [the] fact. When we see ChatGPT produce something in seconds… it is an indicator that it’s not writing,” Warner said.
Warner said the intention and purpose of our words stem from a rhetorical situation, which he described as the context from which the words stem.
“When we write, we write inside a rhetorical situation. This is true, even if we’re writing a diary … There is no rhetorical intention in the large language model,” Warner said.
By contrast, syntax generation by large language models occurs entirely without intention or conscious decision-making.
“It’s like a very big eager dog [that you tell] to go fetch. It’s pattern matching,” he said.
Warner introduced a framework for critically examining the role of AI in our lives with three key components: resist, renew, explore. He urged caution in humanizing AI and placing too much faith in its potential.
“We should resist crediting the hypothetical potential to the present reality … We learn through experience. We live through experience,” Warner said.
“The linearity of writing is illusion after [the] fact. When we see ChatGPT produce something in seconds… it is an indicator that it’s not writing.”
Recalling an experience teaching a student who viewed writing essays like a five-paragraph algorithm, he became aware of the harm of reducing writing to a mechanical process, especially when it comes to self-expression. He also questioned Sal Khan’s claim that AI tutors could improve education.
“What problem is the AI solving?” Warner said. “A teacher [with] 180 essays [to grade] … her problem is she has too many students. So this idea that equity is going to be solved by getting students that can’t afford [it] access to AI tutors [is ridiculous]. I promise you, the rich kids will still have human tutors.”
Warner also underscored the financial burden of expanding AI, citing Sam Altman’s $7 trillion request for OpenAI’s chip costs. This amount is seven times the U.S. education budget from kindergarten to postgraduate, and surpasses the combined annual gross domestic products of the UK and Germany.
“This is unfathomable, double the spending we have on education and management,” Warner said. “So this idea that this thing is our savior, that somehow, because we can’t do this other thing, we need AI, it’s absolutely broken.”
Attendee Vir Patwardhan HM ’28 said he appreciated Warner’s demonstration of the challenges AI poses to education.
“[The talk] was pretty well done,” Patwardhan said. “John [illustrated] some of the fundamental issues with AI in education and why the educational system in the United States needs to transform right away in order to incorporate some of the elements that he spoke about.”
Attendee Jeanne Berrong, a high school English teacher, said she found the talk to be reassuring for the future of her career.
“I feel really validated by this,” Berrong said. “I’ve spent the last two years with a mixture of horror and amazement, watching how many more shortcuts [students] jump through to [save time]. Now I’m hopeful.”
Facebook Comments