Is Critical Thinking Dead?: Dispatches from the History Classroom in the AI Age

By David

Today I want to touch on a topic that is often on my mind when the skies of November grow steely and the trees take on their winter brown: teaching and pedagogy. More to the point, I want to reflect a bit on what I am seeing in my history classroom at Manhattanville University. Are we witnessing the death of critical thinking? Has reading and writing really been replaced by the consumption and production of endless reams of AI-produced slop? Are the kids truly not OK?

This semester, I am teaching a class called “The Historian’s Craft.” For many years our history major did not have a class focused primarily on the historical method. I designed this course to remedy this problem, and it was made a program requirement as part of a significant revamping of the major a couple of years ago. This semester marks the second time I have taught it. I like it. A lot. Enrollment is open and the class attracts freshmen, seniors, and everyone in between. But we tackle some heavy topics including slavery, Reconstruction, the Holocaust, and Israel/Palestine. And the readings, while lighter than what would have been seen on an undergraduate syllabus in prior decades, are not easy.

The second half of the class is built around a significant group project that requires students to build a historical narrative using primary sources from the Manhattanville University archive. They can choose to present their research in any number of ways: documentary film, podcast, interpretive dance, just as long as they are being creative and their project is rooted in evidence.

Students at Manhattanville run the gamut: some could be anywhere, others are barely ready for college. A significant percentage are first-generation college students, and most are here first-and-foremost to improve their job prospects, not to live the life of the mind. This isn’t to say they aren’t curious or eager to learn, but they are, on the whole, transactional: they want to be reassured that what they are doing in the classroom has some broader application or relevance for their career. In other words, one would be forgiven for assuming this student profile, combined with the rapid proliferation of easily accessible AI tools, was especially toxic for humanities teaching and learning. And while I wish I could say that this assumption is entirely misguided, my experiences this semester nonetheless have me feeling cautiously optimistic.

I’ve spoken with several colleagues in recent months about how they’re handling AI use. Some have moved all written reflections to the classroom, where the conditions are more controlled and the temptation to outsource thinking is reduced, going as far as requiring that in-class writing be handwritten. For others, group projects—something I used to outwardly dismiss as a frivolity while quietly suppressing my jealousy of those who assigned them to great success— are increasingly replacing the traditional term paper.

My thoughts on the shift toward “project-based learning” probably deserve a post of their own, but I think it’s a good one. I was genuinely impressed with the projects that came out of the first go-round of this course, which included a detailed map and inventory of the identities of the nuns buried in an on-campus graveyard and a deep dive into the history of the various bars that used to operate at Manhattanville. The university archivist was sufficiently impressed that she asked each of the groups for copies.

Still, I am not quite ready to give up on at-home written reflections. Not yet. Not because I begrudge anyone else the move. After all, reading a bunch of machine-generated facsimiles of thoughtful analysis is dispiriting. It’s simply that I’m not emotionally prepared to abandon the form. At the start of the semester, I told my students that the quality of class discussion would be my barometer for detecting inappropriate AI use. If someone submitted a thoughtful response to the reading but couldn’t reproduce that thinking in class—or in one-on-one conversation during office hours—then I could only assume the work wasn’t their own.

That message seems to have landed. The discussions this semester have been as good as anything I have seen since Covid upended the norms and rhythms of higher education. Not because I think students are foreswearing AI use. I am not that naïve. Yes, I have accused students of submitting machine-generated work, but hardly at levels I would deem apocalyptic. And yes, I suspect that many of them are using AI to digest and summarize the assigned readings. I am of two minds about this. I would prefer they work through dense academic prose on their own. In an ideal world, they would read an assigned work in its entirety, wrestling with the argument through deep, sustained reading (as I discussed in a previous piece, AI can augment the process of slow reading). But to be perfectly honest, my students weren’t doing that before AI. At best, they would read the first three or four pages of an article and hope that I didn’t notice that their written reflection failed to engage with seventy-five percent of the text.

This semester, students were at least grappling with the readings in their entirety, both in their written work and in the class discussion. If they were using machines to help them achieve it, well, I guess I can live with that level of AI use. It beats tearing my hair out as I try manically to police tools that are simultaneously being marketed to students as indispensable. And now that we have moved on to the group project phase of the class, watching the groups really dig into their final projects—unfazed by the fact that they can’t meaningfully outsource the hard thinking work to AI— has me thinking, maybe, just maybe, the kids really are alright.

So what do I take away from my experience this semester in the history classroom? Simply that the situation is not hopeless. Yes, we need to change how we teach. But the truth is, we probably needed to do that anyway. What we can’t do, both as historians and as humanists and social scientists more broadly, is slip into a defensive crouch. To a generation of students who are increasingly transactional in their thinking—no surprise given the cost of higher education—the value and importance of the humanities and social sciences do not speak for themselves (if they ever did). And I realize that for those readers who teach at institutions where students are more…let’s say, academically socialized, some of what I am saying here may not appear immediately relevant. But if there is something I have learned in my thirteen years teaching at a chronically underfunded institution where students lack the sophistication of their peers on more elite campuses, institutions like mine are leading indicators, not lagging ones.

The task before us is to articulate a case for liberal arts education that is neither nostalgic nor apologetic, one that shows why our field and others remain essential to navigating a world shaped by information abundance, technological acceleration, and ethical uncertainty. Students are still capable of the kind of serious intellectual work the humanities demand. AI hasn’t killed that yet. Our job is not just to preserve that work but to make its purpose unmistakably clear.