tl;dr — I just published an opinion article on AI and education in the University of Chicago’s Maroon, where I was a columnist twenty years ago; you can find the article here and, after an introduction loosely tying it to my work on legal writing, below.
The litigator, coder, and self-educator in me all find much to like about AI. I spent two years studying it as a research fellow at UCLA Law, and the release of Claude Code turned my programming hobby into a superpower. I’ve started representing clients in AI-related matters and offering presentations and instructional videos about AI workflows for lawyers.
Part of me, however — and perhaps the greater part — thinks that AI is bad news. It threatens human thinking, learning, and art. It fills social-media feeds, inboxes, and dockets with bad prose. It erodes minds and introduces mistrust into communications whose authenticity can no longer be assumed (Did Dad really write that wedding speech?).
AI is especially likely to hurt education. That’s because structured learning requires a personal investment of time, effort, and care from teachers and students. That effort and those relationships are what enable learning, so they can’t be skipped — but helping users skip such effort, allowing it to be delegated to a machine, is what AI does. AI and education are, in that sense, almost inherently antithetical.
Consider: virtually any nonphysical task that might be assigned in K-12 schools and colleges can now be at least competently performed by the best AI models. Artful prompting isn’t even needed; just upload the problem to Claude Fable or GPT-6 Astra, and it will take care of the rest.
The education world is prone to chasing tech fads. One reason is the fortune that can be made by selling edtech products (and then using them to collect student data); another is America’s permanent state of cultural insecurity about what subjects are worth teaching, and whether schools are preparing kids for the jobs of the future, whatever those might be.
Whether or not AI-enabled automation ever leads to widespread job losses, that prediction alone has triggered calls for sidelining traditional subjects and methods in favor of “preparing kids for the future.” (This can be rephrased as: “changing what we teach to align with someone-in-power’s guess about the future.”) It’s the latest round of a centuries-old curriculum battle.
My essay is a bit off-topic if you follow me for briefwriting tips, but still of interest to lawyers in their capacity as members of the educated public who hope to expand its ranks:
The Death of the Mind
Ryan McCarl
Illustration by Marysabella Llamas for the Chicago Maroon.
If artificial intelligence–powered cheating accelerates America’s educational decline, gullible administrators will share the blame. The University of Chicago, my alma mater, recently arranged for all students and faculty to receive access to Anthropic’s Claude software. It is thus working not to educate students, but to help them bypass education.
AI poses an unprecedented threat to education. Its purpose, evident in the term “artificial intelligence” itself, is to substitute for human thought—to think so its users don’t have to. But thinking is the point of education. Education is the product of the mental struggle for which AI offers itself as a substitute.
Liberal education aims to build the knowledge, skills, and habits of mind that help people solve complex problems and participate intelligently in society. These are built over time, through practice and feedback that gradually change students’ brains. Structured learning activities like essays and problem sets can’t be skipped without losing their benefits. Learning occurs through, and is inseparable from, the investment of time and effort.
AI tools encourage students to skip learning: upload your assignment, the computer will do it for you, and no one will be able to prove that you cheated. Never have students faced a more tempting but self-defeating escape from the woes of homework and grades, or a more addictive and mind-eroding drug.
Although it’s possible for AI to help students learn, that’s the exception. Professors whose expertise lies elsewhere won’t be able to integrate AI in a way that improves their courses; they may know less about the technology than students do. And even if some students have the self-discipline to use AI well, the temptation of using it to avoid unpleasant but edifying tasks is always there, shifting traditional incentives about schoolwork.
The more it seems that everyone is cheating with AI, the more students will cheat, and the more ways they will find to rationalize cheating—which isn’t hard to do when schools welcome the technology that enables it.
Students who prompt their way through school will become dependent on AI. They will not develop the background knowledge or pattern recognition needed to identify complex or nuanced phenomena in the fields they neglect. They will be unable to formulate precise questions when no teacher or textbook is there to supply them.
Far from being useful to future employers, AI-dependent students will be incapable of adding value beyond what the AI itself provides. Their work will often be worse than useless, since their half-baked knowledge and underdeveloped critical-thinking skills will make them pass their misunderstandings on to the model while leaving them unable to guide it or evaluate its output.
AI shortcuts are as tempting and dangerous to faculty as to students; education is a mutual effort that either side can defeat. Great teachers think through their subjects and why students should care. They prioritize learning targets, weigh depth against breadth, and consider how to convey what they know while adjusting to students’ backgrounds and keeping them motivated. Designing useful assignments takes thoughtful effort that AI encourages professors to skip.
Just as teachers have the right to expect that they are providing feedback based on work created by students rather than generated by software, students have the right to expect that an assignment was thoughtfully designed to meet their needs rather than churned out by a computer.
Trust, and the human relationships that depend on it, is at the heart of the AI problem. I want to read prose that the author believes in, not AI slop. I want to mentor law students who write the briefs they sign. When University President Paul Alivisatos explains that the Anthropic partnership will further “our commitment to teaching students how to think independently,” it’s fair to wonder about that statement’s authenticity. Was it generated by the president or by his Claude account? Considering its content, I’m not sure which is more embarrassing.
If universities chase trends rather than safeguard the book-centric activity of thinking, and if they swap their values for the monoculture of efficiency, many will justly wonder what the point of a university is. By embracing a technology that undermines their purpose, universities will help validate the cynical half-truths that higher education is all about signaling, credentialing, and entrenching networks of privilege.
AI hype in all its guises—including Alivisatos telling students that coursework involving AI will “prepare you well to meet the opportunities of a changing world”—invites just this sort of nihilism: What is the point of studying and working if you’ll never measure up to the sameness machine?
The premise of AI, one that ignores how people find purpose in work and suffer from unemployment, is that work should be outsourced to computers at any cost. AI companies’ path to profit lies in convincing executives to pay for tokens instead of employees. What are students to think when their university “partners with” a company that aims to render their careers obsolete and their loans unpayable?
The University of Chicago has always been a home for students who love to learn. Students talk about books and wear shirts extolling the Life of the Mind. It is a former Big Ten school that once bulldozed its football field and replaced it with a library. It now throws the bulldozer into reverse.
Ryan McCarl (A.B. ’08, A.M. ’08, J.D. ’14) is the author of Elegant Legal Writing and a former Maroon Viewpoints columnist. This essay first appeared in the Chicago Maroon on October 6, 2026; an earlier version was published as a letter in the Chicago Sun-Times.
Ryan McCarl is a partner of the business litigation firm Rushing McCarl LLP and author of Elegant Legal Writing (Univ. Cal. Press 2024). For more tips about legal writing and argumentation, subscribe to the Elegant Legal Writing blog and follow Ryan on LinkedIn. McCarl’s book is available on Amazon and Audible.
Subscribe for free to receive future Elegant Legal Writing posts by email:
Connect with Ryan McCarl: Order the Elegant Legal Writing book | LinkedIn | Rushing McCarl LLP | Linktree | Second Stage (blog) | Bluesky | Substack | Twitter (X) | Threads