On my AI policy for my classes
After a number of years swinging wildly between “it’s so over” and “we’re so back” with respect to AI and the classroom, I think I have landed on what seems to me as a stable state. Going forward, in all of my classes, I’m going to take the following approach:
For 100- and 200-level classes, the jig is basically up. I can no longer assume that students are doing their own work for take-home assignments, and these classes are usually too large to implement any of the methods that make deterring said cheating possible. So, perhaps unfortunately, it’s blue books and multiple-choice/short-answer in-class exams for the foreseeable future in these classes. These exams are not without significant pedagogical downsides, and it really is a shame to have someone go through a 100-level philosophy class and not at least begin to learn how to write a philosophy paper. But, when you have 70 or 150 (or more!) students in a class, there just is no other way anymore.
In contrast, for 300- and 400-level classes, I will be doing very little AI-proofing. What I will be doing is starting each semester with an explainer for this policy. I will explain that, if you are in one of these classes, you are either a philosophy major or have taken a number of philosophy classes already. The main thing we should be teaching you, especially in upper-level undergraduate classes in philosophy, is how to think, and specifically how to craft your thinking into an argument in the form of a philosophy paper. So, that’s what we will be doing in the course. I will be assuming that you share this aim with me, and will be engaging with the assignments in good faith. Any AI use you may engage in should be to help you develop this skill, not replace it.
And then, we plug and play. We pretend like it’s 2018 and nothing hurts, and we start writing philosophy papers again. I will further explain to students that, other than the development of their philosophical knowledge and creative thinking skills (both of which require, at the limit, exceedingly minimal AI usage, if any at all), I am offering them nothing else of value in the course. I am not gate-keeping them from an internship at Lockheed-Martin. If they want to cheat their way through the course and I notice they are doing so, I of course will bring it up with the relevant school authorities. But I will not try that hard to do so. Because, in fact, the main things students cheating in these classes will be doing is clowning on themselves. They will be depriving themselves of the actual skills the course is there to teach them, and they will get nothing else out of it. (Even if they pay for the advanced models that can write pretty well and end up getting an A in the class, that A means nothing, over and above the skills they have now deprived themselves of learning.)
Once I’ve said all this, we start doing philosophy, and we (hopefully) never have to talk about AI use in the course again.
There are, of course, still some features of this system that are suboptimal. But it strikes me that they are suboptimalities that just come from trying to teach philosophy in 2026, not from the system itself. I see it as having two advantages over the main alternative AI-proofing at this level, oral exams, which I experimented with in two courses last year. Those exams worked mostly well, but (a) they were massively time sucks (requiring me to give up almost two weeks of instruction time to make time for them, just with two classes of 30 students each), and (b) they didn’t teach my students how to write a philosophy paper. I think getting through a 300-level or up philosophy class and not attempting a serious paper on some philosophical topic is essentially malpractice, and I don’t want to implement that going forward.
Is this belief that students are only clowning on themselves a bit Panglossian? Sure. But I think it is justified. And so I’m going to tell students about it, encourage them to not clown on themselves, then pretend it’s 2018 and nothing hurts.
Analytic philosophy forever. Enjoy the Food Fight.