I closed the classroom door. I said, “This is a safe space. A no-judgement zone.” And I asked my students their thoughts and feelings about AI. They had a lot to say.

As I’m sure you know, the relatively sudden surge of readily available AI tools has upended a lot of things, perhaps most especially education. The bottom line is that teachers, students, and administrators are all scrambling. It’s a mess.

We’re four weeks into the semester now. This past summer, our university unveiled a campus-wide AI policy that exhorts students and faculty to use discernment and tries to lay out some rules about security. Meanwhile, we have purchased a campus-wide subscription to an AI aggregator tool now available to all students, faculty, and staff (in part, from what I understand, as an attempt to effect some boundaries). And then, in the news, all this hand-wringing over the end of humanity, etc., on the one hand and techno-utopian promises on the other.

So I wanted to give my students a chance to talk; I wanted to hear how they’re navigating all this. Over the past couple weeks, I’ve had conversations in two classes, a total of forty students. A small sample size, but revealing. Students are keenly attentive to what’s going on, and they are struggling to figure it all out.

In one class, I started out by presenting the following options for how they might be feeling:

• It’s inevitable. I bow to our AI overlords.
• I am willing to use it for routine stuff like emails, resumes, boring work reports but never anything creative or important
• I already use it a lot and love it
• I already use it a lot and feel guilty
• I resist AI with all my might. Stick it to the man!
• We are living a nightmare. I want to crawl into a cave.

One student immediately reported wanting to crawl into a cave. Who can blame her? Young people are coming into adulthood in a world that has felt on the edge of doom all their lives: climate change, Covid, the American political scene, and now the AI revolution. They have every right to be cynical and anxious. The world is way, way too much.

Another student insisted that “Stick it to the man” was his approach. He understood very well the nefarious greed of the industry’s icons, the environmental harms already underway thanks to the rapid hyperscale data center build-out, and the full-on theft of intellectual property used to train large-language models. He spoke about the sweat shops of underpaid workers tasked with removing the darkest of the dark web from the massive heaps of data that AI draws from—people paid a few pennies to view porn and violence in an attempt to wall it off from AI. This student turns off the AI tools automatically embedded in our search engines and apps these days.

Another student, however, reported using AI and loving it. He spoke about equity: AI gives people with fewer resources the kind of learning tools that other people can pay for. Advice on improving their writing. Testing practice for standardized tests. Help with applications to competitive universities. He uses AI chatbots to explain math concepts the professor didn’t quite get across in class. I’ve heard this from other students, too. They used AI already in high school to explain things they didn’t understand from class lessons.  

The enthusiastic AI-user championed the idea that AI tools can assist and enhance learning for motivated students. For unmotivated students, though, he admitted that AI can actually worsen their lack of motivation. It’s just so easy to take shortcuts and to cheat. For him, though, AI has greatly aided his learning.

This class is a literature course, with a good population of English majors. From them, I heard a lot of worry about what AI is doing to our minds. One student spoke passionately about her worry that AI will hinder mind development in young people. Lots of nods of agreement in the class at this. You might be impressed to know that young people care a lot about their own intellectual development—at least in my experience. They’re also worried about their younger siblings. Will younger kids just adapt to AI before they’ve had a chance to learn how to think without it?

This worried student also wisely pointed out the distinction between generative AI—which will produce original text given a prompt—and what is sometimes called analytical AI, which is used to find patterns in giant heaps of data. I understand that the distinctions are not absolute and the terms vary somewhat, but students do seem to understand that AI is not all one thing, and they worry most about generative AI.

At this point, I shared a bit of my own experience as a professor. I find myself concerned, I said, about the loss of that lonely struggle between your mind and the text. Just you, your brain, all your experiences, in encounter with the words in the book and the minds behind them. And that lonely struggle, too, with the blank page, where you manage to wrestle ideas down—or not.  

Lots of nods. These are people who read and write: they get it.

Am I just being nostalgic? About the good old days when we had to borrow piles of books from the library and write papers by hand on legal pads before typing them up, typos and all, on a typewriter? I don’t think so. I remember how hard it all was. I’m grateful for word processing software, internet search engines, and electronic access to a world of texts.  

But I’m still a fan of the lonely struggle, because that’s how deep, memorable learning happens. That’s how our minds get stronger, more discerning. Even if we engage the lonely struggle with electronic sources and other humans, the best learning still comes from struggle. Just like lifting weights. No pain, no gain: if you want to be stronger, you have to do the reps. Over and over. You can’t program a robot to do your workout for you and then pretend you are stronger.

All that struggle takes time, though. And in a competitive world where efficiency is the ultimate value, time is what students feel they do not have. I understand—and I said this plainly—that students face a dilemma. On the one hand, they understand that the process of doing the work is the point. The product—the research paper, the finished problem set—that’s a trace of the process. A good product is great; you’ve made something valuable. But its value comes at least in good measure from how you got there.

Nevertheless, in education, the product is what students hand in and what gets graded. This puts us all in a bind, professors and students. In a time crunch, with your grade at stake, why not take the most efficient route and just have AI generate something for you? Or at least brainstorm for you? Or suggest an outline? Or summarize a text you can’t take time to read?

So as a professor of English, I now find myself essentially defending difficulty. It’s hard to read difficult texts and grasp them. It’s hard to write. It’s all inherently inefficient. There can be much joy in it—students understand that, too—but that joy is hard won, like a runner’s high. Unfortunately, AI summaries (and those infernal bullet points—ugh!) perpetuate the idea that a text is merely a container from which one can extract some kernels of information (or “powerful themes”). Why bother doing this yourself when AI can extract those kernels for you in seconds? I find myself defending the value of the idiosyncratic encounter: it’s not about the kernels. It’s about the surprises and insights and epiphanies that occur when your unique self tangles up with that unique text.

When I asked students if they feel they are getting mixed messages about AI at our university, they nodded vigorously. Some departments and professors are enthusiastic adopters, even assigning students to use AI tools for particular tasks. Other departments and professors absolutely forbid its use for anything. We professors appreciate that we are given the freedom to manage the AI situation (let’s call it) in ways that seem to us right for each discipline, each course, each assignment. But of course the inevitable result is confusion and frustration.

I asked students if they thought we could achieve a kind of double literacy: able to use AI tools, but also able to do the old-fashioned hard work of reading hard things, poring over data, writing things on your own, thinking and analyzing without the speedy-breezy-greasy results of generative AI. They seemed hopeful about that possibility. Unsure. Hopeful. Unsure.

Students expressed appreciation for the professors (including me, as best I can) who are trying to adjust our teaching practices and assessments to demonstrate that we value the learning process and not just the product. One student remarked that she rather likes spending time journaling by hand, for instance. Many of the students expressed appreciation for the idea of creating AI-free zones where they could revel in reading and creativity.

In fact, my other class, full of students committed to the arts in all forms, expressed deep worry about what AI is doing to human creativity. What will happen to creative industries like graphic design? Will human creativity become a swamp of AI-generated slop? How can they continue to perform music, create sculpture, produce films, write good stories, and do photography if we shunt all that off to machines? Is AI really going to take all our jobs?

I’d like to believe that in fact, human-generated creative work will increase in value. And that humans will always value the tactile, the real, the genuine encounter with each other and the world. Maybe more so now? I wonder. I can’t be sure.  

You might be wondering what my policies are. In my literature and writing classes, I clearly ask students not to use AI for anything. Not for informal or formal writing. Not for summarizing texts for them. Have I caught students doing it anyway? Yes. Quite a few times. Always the explanation was “I was crunched for time” or “I just used it to give me an outline.” I’ve had to come up with specific ways to enact reparative justice in these instances. There has to be a cost in grade reduction but I also want to give them a chance to do it again, genuinely, maybe with some tailored adjustments to the assignment.

Have students “gotten away” with using AI in my classes? Oh of course. I can’t chase down every suspicion. And as Katherine Rundell points out in her excellent essay on AI and student learning, it’s easy to catch students using AI badly. If they’re using it well, not so easy.

For me, the worst part of this whole mess is the erosion of trust. One of my students lamented that in high school, she noticed this, too. Teachers had to be suspicious all the time. I thanked her for perceiving this, because that is what I lament the most. Over my many years of teaching writing, the greatest joy has been encouraging students whenever they write well. When I spot good writing, whether just a sentence or a whole masterful essay, the best part of my job is pouring on the praise. Now, however, I always wonder: Is this actually her work? Or is this just AI? I find my praise more muted because… what if?

There’s so much more to say. For the moment, I simply report on my small, slapdash study of student attitudes at one university. Students are keenly aware of the industry’s severe and myriad ethical violations. They are keenly aware of the environmental threats. In fact, some of them wonder why on earth our university’s policy only passingly mentions the obvious moral offenses of the industry. Some are furious that the university has spent money on that AI aggregator. Students perceive how this AI revolution (or whatever) is embedded in vast geopolitical and economic dynamics, now affecting them directly.

They also understand that there are good and bad uses of AI tools. They get the difference between analytical tools in the service of advancing knowledge toward good ends and generative AI that might be helpful with certain tasks but can also produce slop and deception. AI is giving them yet another reason to wonder what their future holds. Even in a competitive, fast-moving, results-focused world, many students remain genuinely committed to their own learning. Others, inevitably, are succumbing to AI’s allure and using it to cheat, short-circuiting their learning.  

My conversations with these students ended with no satisfying conclusions. We are all struggling. All I can say is: please be kind to young people these days. They are coming of age in a swirling, confusing, difficult world.

I’ll leave you with an excerpt from a bracing poem written by one of my students, which she generously permitted me to share with you:

When I write I think a lot.
I write a little, but mostly I think,
And in my mind I stew over life, and the machine, and my role in it.
I write a little, and I think
that if the internet gets to splice
and butcher my great love then
maybe I should treat it with the mercy it offers me and mine—
get a kill switch, shut it down, pull the plug because nobody else will and I’m running out of
creative solutions that involve me being kind.

(Excerpt from “I Rage Against the Machine” by Cora Visser. Cora won her age division in the Dyer-Ives poetry competition with this poem. )

A good essay in the New York Times about how universities are coping.

A helpful overview of the environmental threats AI poses. Another helpful overview, focused on data centers and the backlash against them.

One of more infuriating aspects of the industry right now: sweetheart deals with fossil fuels and the use of AI to increase fossil fuel production.

A study (a small one) showing that students are worried about their own learning.

A beautiful sermon from Rabbi Zvika Krieger on the spirituality of AI (in response partly to Pope Leo XIV’s recent encyclical and the spiritual implications of certainty and uncertainty. Thanks to my friend Ace Leveen for sharing this with me.

Further notes:

Calvin University is hosting a conference on AI in October, in an effort to work through some of these issues more deeply.

Students described here gave me permission to share the things they said. I am grateful for their honesty and trust.

I never use AI to for any stage of the writing process. In fact, my recent book contract specifically stipulated that I could not use AI to write anything in the book. Of course, search engines now have built-in AI summaries, which I do use, with skepticism and caution (and some annoyance at all those stupid bullet points). There are other types of AI tools (voice-to-text, text completion) that I use. I’m not a no-AI purist. Just very, very cautious. And very aware of the many moral failings of the industry.