Embrace the Use of AI in Student Work

Generative artificial intelligence (AI) tools like Grok, ChatGPT, Claude, et cetera will radically change student assignments and evaluation in college classes. I have already explained how AI means that students can no longer be forced to do the reading. How will college professors change in response? Consider some likely approaches.

They can ignore AI because it does not really affect their approach to student evaluation. If you are teaching calculus or any course in which the primary grading tools are in-class exams, then AI may not matter. You still lecture. You still give students homework assignments. All those assignments are now trivial to do with AI, and there is no way for you to prevent its use. But, as long as you grade on the basis of in-class exams, students will have every incentive to do the homework honestly, if only to practice for the exam itself.

Faculty can ignore the change and be amazed at how much smarter their students have become. Any traditional out-of-class assignment—the most common of which is the essay—is now easy to complete at the highest levels of quality with AI. (It is astounding how many professors seem to be in denial of this basic fact.)

[RELATED: A Faculty Guide to AI Pedagogy and a Socratic Experiment]

They can fight AI by banning its use by their students. This is hopeless. Some institutions—blessed with engaged students, a serious honor code, and a tradition of take-home exams—might be able to hold off for a while. Eventually, however, the competitive pressure will just become too much. AI is just too good at writing excellent college essays on any topic.

Faculty can “go medieval” by requiring students to complete all written work in class. Some institutions, like the University of Austin, are moving in this direction. This can work! But the costs are high.

Faculty do not want to read handwritten essays, especially as student handwriting deteriorates in our technological age. No one wants to sacrifice classroom time to evaluation exercises. One approach would be to emulate the typical law school and have a single monitored writing exercise during exam time as the main grading tool. This avoids wasting class time. The shared use of a dedicated room with monitored computers can also solve the handwriting problem. The cost is high for a single evaluation opportunity. Such pressure is an accepted part of the law school experience.

Would it work at the college level?

Faculty can embrace AI, encouraging students to use it in all of their assignments. I recommend this approach. We should no more forbid the use of AI than we do the use of calculators or spell-checkers. (There is a case in K -12 education for teaching the “fundamentals” of unassisted mathematics and spelling. But that argument hardly applies to college students, at least at elite schools.)

How can instructors embrace AI?

Begin by using AI yourself. How would Grok answer your favorite essay prompt? How accurate are the references suggested by Claude? How good are the theses statements created by ChatGPT? Generative AI is the future of education and scholarship. Use it or be left behind.

Be explicit about your assignments’ goals and the time you expect students to devote to completing them. Assigning a traditional 2,000-word essay with a standard prompt is bad since AI will provide an A-quality answer immediately. You presumably want students to spend more than a few minutes on your assignments. You want students to wrestle with the themes of the course and its readings. You want them to create something new.

Scale assignment difficulty to account for the power of AI. If you want students to spend 10 hours, you need to design the requirements so that a typical student needs 10 hours to complete the assignment. The best approach is to require more. Instead of a 2,000-word essay, require an illustrated 2,000-word essay. Instead of still images, require videos. Put in on the web. Add a voiceover. And so on. Although the AI tools that allow any undergraduate to create a Bill Moyers quality documentary are not available yet, they aren’t too many years away.

[RELATED: The Use and Misuse of AI in Higher Education Writing Courses]

Is this really possible for average undergraduates?

Yes! Consider the Great Books course I taught last fall. I gave students this example of a 2,000-word essay with some simple videos created with AI. They, in turn, produced some excellent final projects. They spent about as much time on these projects as they would have on an essay-only assignment in the era before AI. They wrestled with our themes and readings. They created original works. What more can any professor ask?

The main causal effect of your class on a student is the difference between the quality of the work that the student produces after taking your class and the quality of the work that she would have produced if she did not take your class. Generative AI does not change this fundamental reality. But its existence means that our traditional approaches, especially out-of-class essays, work much less. Outside of heroic measures, there is no avoiding AI. As teachers, we might as well embrace it.


Image of ChatGPT homescreen by Emiliano Vittoriosi on Unsplash

Author

  • David Kane

    David Kane is the former Preceptor in Statistical Methods and Mathematics in the Department of Government at Harvard University.

    View all posts

Leave a Reply

Your email address will not be published. Required fields are marked *