My first significant interaction with ChatGPT occurred about two years ago, when I asked it to generate a modern worship song based on the first seven lines of the “Aeneid,” and it spat out a chorus about Aeneas beginning “But he had faith / He had hope / He knew he had a higher goal.” 

After I finished crying, I decided that generative AI was better left out of discussions of classics or liturgy. Freshman orientation at Hillsdale only served to confirm this conviction: the English professors made it quite clear that AI was to have no role in my writing, whether in brainstorming or in the process of essay composition itself. 

As a student of classical languages and literature, I more or less concur with the English department. I personally don’t have a use for ChatGPT, as I’m perfectly capable of doing my own translations and composition, and I am rather bored by the fact that you can coax an AI into generating senseless but amusing responses to leading questions. 

Recently, however, I had a conversation about AI with Ian Church, an associate professor of philosophy. He pointed out that generative AI is now a part of our world and it’s not going away any time soon. Thus, for Church, it’s worth investigating precisely how one might go about using ChatGPT as a tool in the classroom. 

He informed me of a recent article by Justin Weinberg about a philosophy professor at the University of St. Thomas in Minnesota who proposes using generative AI as an “interactive course tutor” to check students’ responses to homework assignments, as a tool for producing new assignments and homework, and as an always-available resource for answering students’ questions about practical aspects of his courses. 

Church also proposed a number of other potential uses for generative AI — it could play the research assistant, act as a tutor, or even serve as a sounding board off of which to bounce ideas. 

All in all, AI has the potential to be a positive good in our classroom, English department moratoriums notwithstanding. That said, I’m not completely convinced that AI should be used in humanities classrooms, no matter how useful it might be. Are the humanities antithetical to unhuman AIs? In considering the question, I’d like to present an analogy.

Imagine, if you will, the first time people began writing down epics, say Homer’s “Iliad,” circa 750 B.C. Prior to this, poetry had been exclusively spoken, making human memory a critical element of the performance of epic. Now, this new-fangled writing will do away with the need for a prodigious memory. Who needs to memorize nearly 16,000 lines when you can simply read them aloud? But by losing the memorization component, are we running the risk of throwing away a fundamentally human aspect of ourselves?

To put these questions another way, why brainstorm yourself or write an essay when you can get a bot to do it for you? Are we willing to risk the atrophy of the ability to write well, a fundamentally human endeavor? 

In some ways, my hypothetical writing naysayers in the 8th century B.C. were probably correct: writing does relieve the average person of the daily necessity of memorization by dint of its permanence. I, for one, do not feel the need to memorize things I know I can find later in a book. (For instance, I didn’t memorize the name of the book in which I first came across the idea of writing being potentially adversarial to memory, although I have a sneaking suspicion that it was Nicholas Carr’s “The Shallows.”) 

But, in more important ways, the imaginary opponents of writing were wrong. The existence of writing has not robbed the human memory of its significance. Instead, paradoxically, it has become a means by which memory is rendered nearly permanent. If we follow the logic of the analogy, perhaps we can hope for unprecedented benefits from AI as well. 

Questions will remain regardless of AI’s ultimate impact. Is AI good for our classrooms? Can we reconcile human creativity and logos-less logic? And could this essay  — yes, the opinion piece you’re reading right now  — have been generated by an AI? I’ll leave the answers to those algor—er, writers more eloquent than I.