In February, the Maroon asked the University community to share their thoughts on the increasing prevalence of generative artificial intelligence tools across UChicago. Here’s what our readers had to say.

To the Editor:

Large language models are incredible tools for processing and analyzing vast quantities of data. If used responsibly, they are valid and game-changing. That being said, I hate the way I see these models used in the classroom. Actually, “hate” is not a strong enough word.

If you want to cheat on a test, that doesn’t bother me—that’s your experience you’re robbing yourself of, probably for an understandable reason. However, if you use generative artificial intelligence to do your thinking for you and to cheat on papers, that poisons my educational experience, robbing me of the intellectual environment I want to have at this school. Furthermore, it poisons the professor’s experience of teaching classes in their field.

Last quarter, I had the privilege of taking a class with one of the greatest minds in the field of philosophy. Watching him explain that he could not count our midterm papers for a grade because too many people had used AI was painful and infuriating. It was heartbreaking to see this man, who has dedicated his life to the thing he loves, face the fact that so many people taking his class couldn’t be bothered to care. This also robbed me, as I had spent long hours writing an essay worthy of this class, which had to be discarded because too many people had spent too long on Instagram Reels.

If you want AI to do your thinking, writing, and arguing for you, you do not belong at this university, or any university for that matter. I went to college to learn from academics and my peers. If I had wanted to argue with soulless robots all day, I would be at home, bedrotting and asking Grok if Clavicular got frame-mogged at Arizona State University.

Charles Robinson is a third-year in the College.

To the Editor:

On Sidechat, I saw a poll titled “Do you use chat/gemini/claude/etc on psets (be honest)?” 183 students voted: 149 “yes,” and 34 “no.” This tracks with what I’ve seen during my three years at UChicago: I’ve seen ChatGPT open on a laptop every day I’ve been on campus; I’ve heard about students opening large language models (LLMs) in discussion sections and speaking aloud AI-generated points verbatim; and, of course, I’ve noticed that using AI for p-sets is common, if not the norm, across majors. I’ve also had humanities courses where entire paper assignments were made void for the entire class because too many students were obviously using LLMs. So, how is AI being used in the classroom? Illicitly, by almost everyone, for everything they can get away with. And our professors and administration are willing to let students get away with a whole lot.

The purpose of generative AI is to avoid doing work. If the work in question is the substance of a class, allowing generative AI use is no less of a pedagogical atrocity than giving elementary schoolers calculators while they’re learning arithmetic and shouldn’t be expected to result in a different result. Consequently, I don’t think it’s hyperbole to wonder whether—if generative AI use is not excised from the university system—in four or five years the average “graduate” of an “elite” college like UChicago will be an utter moron distinguishable from other college graduates only by their capacity to reach the low ceiling of a standardized test and get some tears out of an admissions officer.

Yet, given the obvious gravity of this situation, our administration has done shockingly little. The responsibility to deal with AI use rests on them, not on student self-policing or “honor codes.” Students do not generally come to college expecting to pass through trickier and trickier ethical dilemmas resembling something out of a moral philosophy paper—they expect that hard work and understanding will be rewarded, and that the opposite will be punished. If this is not the situation, how can our school be expected to function, and how can we expect students not to defect given the ridiculously skewed payoff matrix? The students that are occasionally caught are not scapegoats, but their existence does give lazy administrators and professors an excuse to look away from the problem and congratulate themselves on the official integration of AI in the classroom. They are presumably unaware or willfully ignorant of its unofficial dominance in their classroom, which makes these initiatives completely superfluous. The way the system works right now is equivalent to a professor entering his classroom during an exam, failing the first person he sees cheating, leaving, and calling it a day.

Universities have spent billions of dollars outfitting their classrooms with the latest technology. How hard would it be to just go back to the basics of blue book tests, oral examinations, or in-class essays? Ultimately, these recommendations, since they’re so obvious, are probably pointless. Our administration is often incompetent but rarely stupid, so they are likely taking piecemeal measures not because they don’t know what to do to restore academic integrity, but because they don’t really want to do it.

Why? Because, despite the endless platitudes, they must not be sure what the purpose of a university education is. If college is about just picking an interesting-sounding major, learning whatever you feel like, and then going to work in consulting or finance, then generative AI is perfectly compatible with college. And if college is about learning marketable skills you can use in your job, say goodbye to philosophy, English, Germanic studies, and all the rest. For some time now, the idea that a liberal arts education should be focused on moral development, cultivating taste, and broadening students’ horizons has been something of a joke given the entanglement of school and market. Generative AI use is just the coup de grâce. Until we recover some shared understanding of what “college” should be, we should be content only to try and assuage the continued and inevitable decline of the university system to the best of our abilities. We cannot “defend” the university against AI until we confidently know what the university is supposed to be.

Owen Yingling is a third-year in the College.

To the Editor:

A few months ago, another campus publication put out a cartoon that appeared to enact the unmasking of an unsavory truth: behind the curtain of what we call “higher education,” students at UChicago are using ChatGPT to write their essays, and our professors are using ChatGPT to grade them. All parties feel the sting of regret, but the buzz of modern life moves quickly enough to equate the concessions. The ironic figure of man central to this picture is one which scoffs and assumes that all must be as pragmatic, as self-serving, and as empty inside as he is. There is a neat symmetry, such an image would suggest, in the corners you might cut in your classwork and in the overarching ways in which man, as a whole, is being diminished—getting more comfortable, more mediocre, and more indifferent as we get “better” at using these machines.

This idea of large language models’ (LLMs) inevitable pervasiveness, however, is absolutely not the case, and to imply that it is the case is an affront to the good-faith efforts that stand against it. There is work that is fit for LLMs to do—this is the drudgery of tasks like adding periods and commas to an interview transcript or copying a list of names from a photograph to a spreadsheet. Just reading these instructions makes the eyes glaze over; these tasks do not merit the attention of intellectuals with training.

LLMs and the sanitized, chipper voices in which they speak ought not be used for any writing task intended to deceive another person into thinking they are reading human writing. In real life, machines are placed within networks of relationships and of trust, and our classes here contain no such legitimate use cases. Giving up on learning attempts to cheapen what cannot be cheapened—the value of a real education, which we get at the University of Chicago as we do in few other places. Within a college, as students learning from professors, we have nothing without the trust that this process of learning is simply too important to cede. After all, there is no greater gift than taking possession of oneself through education, and this is a selective school. As long as LLMs do your work, you are a wasted seat.

Josie Barboriak is a fourth-year in the College.