Blue Books and Oral Exams Are Not The Answer
Here’s how we should respond to generative AI's impact on writing instead
Thank you for being here. As always, these essays are free and publicly available without a paywall. If you can, please consider supporting my writing by becoming a patron via a paid subscription.
Imagine getting rid of marathons because of the motorcycle. Generative AI is prompting (pun intended) many educators to rethink how they teach writing in their classes. This is good. However, not all responses are created equal. Today I want to explain why moving back to blue books and oral exams is not the panacea that it may seem and make a case for a better way to respond.
The present reality is that students have generative AI at their fingertips and can chat with an LLM just as easily as they can message their friends. And it's not just on their phones. Google’s “Help Me Write” feature is baked into the right-click menu in Chrome for crying out loud. The development and widespread integration is significant enough that it warrants our consideration as educators, especially as we think about teaching them to write.
Out of the desire to protect the learning in their classes, instructors are understandably concerned that having an LLM lurking over the shoulders of students as they write will be an issue. I agree. But several threads that I've read over the past few days have reiterated a theme that I've heard for some time now. It goes something like this: now that students have easy access to LLMs, we need to adjust our assessment to make it harder for them to use them by moving our out-of-class writing assignments and research papers to in-class essays or oral exams.
Listen, LLMs, especially in the hands of students who are still learning how to think, reason, and write are full of potential pitfalls. Nevertheless, how we navigate this new minefield matters.
I am not opposed to in-class writing assignments or oral exams. There are pedagogically justifiable reasons for both. But using them just to make it harder to cheat is not one of those reasons. This sort of change says more about the way we see our students and, ultimately, illustrates that we fundamentally don’t believe that we can trust them to engage in the hard work of writing without succumbing to the temptation of outsourcing their work to an LLM.
Getting rid of long-form writing assignments in your class because of generative AI is like getting rid of the marathon because of the invention of the motorcycle. Instead of spending time building barriers to try and prevent our students from cheating, we would be much better off spending our time revising our assignments to align with our desired learning outcomes and communicating that rationale more clearly. Whatever time we would spend building barriers would be much better spent building bridges.
What, after all, are we trying to do as educators? Yes, we're trying to build skills. But we need to think carefully about whether we're using training wheels or scoot bike pedagogy. As we teach, we need to look beyond the immediate skills that we want our students to acquire to the type of people we want them to become. To do that, we need to help them to build character.
Writing and Running
Before we talk more about the issue at hand, I want to take a slight detour to explore the marathon analogy a bit more deeply. It’s been useful to me this last week as I’ve been wrestling with why the responses I’ve been seeing online haven’t been sitting well with me.
If you have any familiarity with running, the name Eliud Kipchoge likely rings a bell. Widely considered one of the best marathon runners of all time, Kipchoge won the Olympic men's marathon in 2016 and 2020 and is planning to try and three-peat in the next few weeks at the 2024 Summer Olympics in Paris.
Kipchoge is fast. Like 13.1 miles per hour fast. I’m out of breath just typing that. If you want to cover 26.2 miles on foot as fast as possible, Kipchoge is your guy. But of course, if you only care about covering 26.2 miles as quickly as possible, then doing it on foot is highly suboptimal. There are many other ways to cover the ground more efficiently.
And yet, despite these much more efficient options, we still run marathons. Why?
There are lots of things in life where the fastest, most efficient way is not the best way. In fact, this is true of almost all of the most meaningful things in life. In the case of running marathons, we are not celebrating what can be done by eliminating human limitations but by pushing up against them. Kipchoge is extraordinary because he demonstrates what is possible by embracing and pushing his limits—in addition to some healthy genetic predisposition—to demonstrate the extraordinary levels of endurance and speed achievable by a human.
Align the training and the motivation with the goal
Consider telling Mr. Kipchoge that instead of running the marathon in the 2024 Olympics, he would be competing in the 100-meter dash. Despite being the best marathoner on the planet, he would get smoked. Why?
There are several reasons, but at a fundamental level, he wouldn’t be competitive because the marathon and the 100-meter dash are designed to highlight very different aspects of a runner's performance. While the marathon is designed to test long-distance speed and endurance, the 100-meter dash focuses on short durations of maximum speed.
In helping students to think and engage with the writing we’re asking them to do in our courses, there's nothing wrong with long-form essays, in-class writing assignments, or oral exams. They each have their merits. But they are no more interchangeable than the marathon and the 100-meter dash.
You cannot simply swap a long-form essay for an in-class essay without rethinking the learning goals for the assignment and realizing the strengths and weaknesses of each format. A long-form essay completed outside of class time is designed to give students space to wrestle with an issue over time—to consider various perspectives, incorporate a variety of sources, and carefully articulate nuance. Long-form essays lend themselves to revision, allowing students to return repeatedly to their work to reassess and strengthen it.
In contrast, an in-class writing assignment or oral exam is the 100-meter dash to the long-form essay's marathon. It's a sprint, evaluating a student's ability to synthesize information quickly, on the spot, without time to let the thoughts percolate and revisit them later. There is nothing wrong with training this skill. But it's not working the same intellectual muscles as the out-of-class essay.
There is a better way
Listen, I understand the consternation surrounding generative AI. I’m concerned about the impact it will have not only on students' ability to write but also on their understanding of what writing is worth in the first place. Unfortunately, whether we like it or not, our students have all been given the option to ride the generative AI motorcycle instead of running.
But all is not lost. Just like we still run marathons and celebrate the athletes who run them, we must resist the urge to hastily toss aside our existing writing pedagogy. Just like running a marathon still has value even though you can cover the ground faster in other ways, writing essays is still deeply valuable even though you can get your favorite LLM to extrude a bunch of syntactically flawless text with a short prompt.
I am not arguing that we stick our heads in the sand or that we leave our assignments unchanged. We certainly will need to adapt our pedagogy to respond to generative AI. But whatever choices we make, they must be grounded in thoughtful motivation and defended with pedagogically sound reasons. The tasks we ask our students to do should be specifically designed to train them in a certain way.
Instead of scrapping our existing assignments, we should ask deeper questions. Questions like:
What is at the root of our worry about students cheating?
How might we better help our students to understand what our courses and assignments are designed to do?
How might we better motivate our students to choose the engaged path and avoid the shortcuts that will rob them of learning?
How might we cultivate a classroom community of trust, honesty, integrity, and character?
If you can pivot your long-form essay to an in-class writing assignment without substantially changing it, I'd question whether it was a well-designed assessment in the first place. Instead of throwing good money after bad, we should get back to the fundamentals and think about how we want our students to be formed. Only after we figure that out can we make sensible and wise decisions about the types of assignments they should be working on.
If you’ve got something to say, please leave a comment. What resonates with you? What am I missing?
I’m also particularly interested to hear from those of you who teach writing. How are you thinking about approaching your courses this fall?
Please also consider sharing this with colleagues who may get something out of it or restacking it!
Last thing: if you’ve got something else to say or just want to say hey, send me a message.
Reading Recommendations
Last week I shared part one of
’s reflections and predictions about generative AI this fall. He recently published the second part as well.relaunched his Substack this week with a post about the fruitlessness of the right systems with the wrong goal. Seems like lots of themes that are valuable for our conversations around generative AI as well.This fall will not be an easy term. We haven’t had one of those in a while. Yet, if we find the capacity to act, this is a moment when we can still shape how our new tools are to be used in our classrooms. If we’re too busy, if this semester is too hard, then the moment will pass. The tools our students are using to shape their world are starting to shape our institutions. That’s how new tools change society, gradually and with little awareness of what we are creating with them.
writes a teardown of season 3 of The Bear that I never could—because I couldn’t even get through episode 2.So what I realized I needed were two things. First, I needed a coherent picture of what matters to me, and how those meaningful things instantiate themselves on different time scales: Lifetime, 3-5 years out, 1-2 years out, in the next quarter, the next month, the next week. Second, I needed to install habits where my Clarify process is inseparable from intimate knowledge of this big picture.
The Book Nook
Amid summer travel and work my reading rhythm has slowed down a bit. Still in the middle of a few different books, but recently picked up a collection of Flannery O’Connor’s short stories which I am enjoying. I particularly enjoyed reading the foreword which gave some context about Flannery’s story that I was previously unaware of.
The Professor Is In
Had a great time last week in New York City spending the day with Praxis. In addition to some great ideation time in the morning, I was glad to have a chance to participate in the first part of the design session they ran in the afternoon.
Leisure Line
Got out on Friday night to see Needtobreathe live at the Toyota Oakdale Theatre in Wallingford, CT. Just a great show from these guys, as always.
Still Life
The best ice cream in the world is found at a little place called Rich Farm Ice Cream in Oxford, CT. Whenever I am in CT I do my best to hit three food stops: Rich Farm Ice Cream, Pepe’s, and Sally’s.
Well-timed essay for those of us retooling courses for this fall. I decided last spring that I was going to preserve a long-form research essay assignment in the graduate history of higher education class I will be teaching. I am redesigning the class with more structured group activities during class time and invite the students to experiment with an LLM as an educational tool that can give them feedback on their writing along with peer review and my own reading of their drafts.
I've taught the course three times online, and this is my first in-person teaching since the pandemic. The return to a physical learning space is allowing me push hard into my Deweyan beliefs in student-centered social activities as the path to learning. Teaching undergraduates in the past using those methods, I often received feedback that the approach was uncomfortable and too difficult, something that the literature on this approach bears out....many students don't like it. Online really didn't work either because I couldn't get past how easy it is for students to disconnect when on a screen.
I'm hoping that the shared questions and confusion about the educational value of generative AI tools and the fact my students are mostly master's-level enrolled in a professional higher ed administration program, will mean they will be more open to experiment as well as student-developed and led assignments. The class meets on Friday afternoons, which has limited enrollment (so far 7 students when usually I get 20), which may also help create the dynamic I'm looking for. Plus, I'll get a teaching space suited to in-class activities, which has been a challenge for me in the past. Nothing like fixed seating to add one more challenge to active learning.
Your essay, along with some of Marc Watkins recent essays are going on the first day syllabus for students to better understand where I am coming from when it comes to writing. Thanks, as always, for such a clear articulation of an important educational idea.
Learning long-form essay writing in high school played a critical role in preparing me for research. Sure, I could have easily cheated and re-wrote other people's work just as easy as talking to an LLM, but it wouldn't have helped me reach my goals. I agree that in the end, focusing on creating assignments torwards learning goals rather than to block cheating is the way to go.
Also, Needtobreath is my favorite band, such good performances and Christ-centered messages.