Thoughtful

I remember struggling with Hannah Arendt the first time I read her. She used the word “thoughtful” in a way that native English speakers typically don’t: She used it to mean considered, as opposed to the more common meaning of considerate. Where I might refer to a thoughtful gesture, meaning that it was considerate, she would refer to a thoughtful approach to politics. The latter wasn’t about manners or etiquette; it was about thinking through what we were doing, or trying to do. It was about being full of thought, in the sense of having engaged in serious (and collective) reflection. The absence of reflection, or what she called banality, paved the way for tremendous evil.

With more miles on my personal odometer now than I had then, I see her point. Recently I’ve come across a couple of particularly thoughtful pieces, in Arendt’s sense of the word, on how to respond to AI in higher education. They aren’t as banal as either “go away and leave me alone” or “now we don’t need education at all.” Instead, they offer examples of what serious and collective reflection could look like, each with an implicit promise of helping colleges and universities figure out more clearly what it is they should be doing. I commend them to my wise and worldly readers, and to anyone else looking for hope.

The first is the widely noted report from the Massachusetts Institute of Technology on “AI Use in Teaching, Learning, and Research Training.” It’s a striking document in any number of ways. The most glaring, by comparison to most of what I see in my role, is the centrality of learning, as opposed to job placement.

Some of that may reflect a sort of privilege of rank. MIT can safely assume that its students will get jobs. Its admissions policy and its curriculum are famously exacting enough that people assume anyone with a degree from there must be pretty sharp. When you prescreen out anyone who hasn’t shown a track record of being extraordinarily good at doing school, you can elide certain questions and get away with it. Must be nice. They also don’t have to worry about transferability, which opens up options for grading that don’t exist in my world.

But even granting that, I was heartened by how thoroughly and candidly the report addressed the challenge of AI. At one point, it scolds faculty who have replaced research assistants with AI. The argument there is that yes, sometimes AI can complete certain tasks more quickly, but that defeats the purpose of the institution. Research assistants are future researchers, and part of the point of assistantships is learning how to do research. Replacing entry-level researchers amounts to eating the field’s seed corn.

The report offers eight principles, one of which (“intentionality”) is a rephrased version of Arendt’s thoughtfulness. Another, “no one size fits all,” offers a refreshingly practical suggestion for AI policies: Rather than either a single collegewide statement or leaving every professor to their own devices, offer a short menu of options. In effect, they’re suggesting that the role of leadership is to devise the right multiple-choice question, and to allow faculty to pick the answer that make sense for a given course or task. It’s the kind of retrospectively obvious solution that reflects serious reflection in the process of its development. In other words, they make it look easy. I mean that as a compliment.

The second is an extraordinary post by Christine Nowik, an English professor at Harrisburg Area Community College in Pennsylvania. In a wonderfully practical way—complete with a workbook!—Nowik suggests taking the challenge of AI as an opportunity to reflect collectively on just what it is that we’re trying to do. She suggests replacing questions about AI detection and policy compliance with questions about what we want students to learn and where “friction” is an indispensable part of learning.

Nowik’s point that the task of leadership now is “not to eliminate complexity but to organize it.” That’s consistent with MIT’s approach of multiple-choice policy options. I’m thinking I may get “not to eliminate complexity but to organize it” embroidered on a pillow.

Some of the principles she lists reflect the realities of underfunded, teaching-intensive institutions. In other words, they’re applicable to institutions that aren’t MIT or its peers. “Do not create an innovation penalty” and “let some old things die” are far harder than they sound, but all the more important for that. Nowik is keenly aware that institutional mandates often create new work for faculty (and, I’d add, everyone else on campus) when they aren’t paired with deletions or subtractions. When base teaching loads are 5/5, as they are here, and entire offices consist of a single person, awareness of burden is key. Nowik suggests creating what she calls “slack” to give people time and resources with which to experiment, and to have enough perspective to know that some experiments fail. That’s why they’re experiments.

At this point, I’ll make a bit of her subtext into text. Creating the space to experiment (and occasionally fail) requires money. Decades of public-sector austerity have removed most of the budgetary wiggle room that might once have allowed for more ambitious experimentation. Current political trends being what they are, we may need to look to philanthropy (as opposed to legislators) to find the resources to make some of these ideas possible. We need look no further than last Friday’s story about Ohio state Senator Jerry Cirino and his dismissal of shared governance as “garbage” to understand just how important, and difficult, developing a truly thoughtful approach to the challenge of AI will be.

For all of her insight into thoughtfulness, Arendt was notoriously indifferent to economics. (Her friend Mary McCarthy once challenged her directly: If you remove “the economic” from “the political,” what’s left?) We can’t be. Our resources are limited, and in some cases, the people who control access to those resources have taken impulsive ignorance as a sign of resolve. We need resolve in the service of wrestling with serious questions. Kudos to both MIT and Professor Nowik for showing how it can be done.