Many of the conversations I’ve seen about Artificial Intelligence (AI) center on how students might use or misuse this powerful tool. I’ve had my share of conversations about how to AI-proof written assignments, and explored some ways we might lean our activities more directly into the weaknesses of chatbots. Many of us have received university-level guidance which helps us know how to respond to AI-assisted writing when we see it (or when an online checker tells us to take a deeper look).
That’s all good, but it also feels fairly reactive. And it misses another consideration.
In “Human Meets AI: Helping Educators Navigate Their Emotions About Technological Change,” Dr. Leo S. Lo writes, “the capabilities of generative AI tools might feel like an unwelcome incursion into the hallowed realm of the personal, human-centric field of education, stirring up deep-seated existential questions about the future role of teachers and administrators.”
If AI can write a student’s paper, or handle their responsibilities on the discussion board, what else can it do? Could an AI chatbot be integrated into an online grading platform and provide feedback to students? Could it interact on the discussion board? Could AI diminish or wholly supplant the role of teachers in online education?
Experiments in automation
In Futureproof, Kevin Roose, a New York Times writer, talks about the ways AI can do things quickly and on-demand. He delves into the disruptions that AI, and technological automation in general, can have on society.
It got me thinking about how AI might interact in the online classroom. I copied a recent student discussion response from one my classes and anonymized it. Then, I dropped it into a ChatGPT window and typed, “Can you give me feedback on this?”
The instantaneous feedback it spit out was remarkably good. It was encouraging and incisive. It reframed what the student said and offered suggestions for improvement – in a bulleted list, no less!
I repeated this experiment with a student essay and asked ChatGPT specific questions based on several defined rubric categories. And, again, the feedback was sharp, thorough, and targeted. While the feedback might have occasionally skewed toward the generic, it was always clear and, in my eyes, accurate.
Would students know the difference? Would administrators increasingly obsessed with efficiency and trackable metrics care?
Roose’s Futureproof also outlines a few things that AI cannot do, and a lot of it centers around the human domain of offering support, encouragement, and emotion. Additionally, ChatGPT doesn’t deal well with unexpected issues. In fact, before we start getting too worried, I should note that Roose (and others) have identified numerous things that AI isn’t very good at.
The good news is that certain areas where AI struggles aligns with much of what teaching is centered around. It’s about humans being human in the classroom, doing the complex, emotional, and messy labor that (as of right now) only we can do. My worry is that as we increasingly seek to define a complex and interactive concept like “engagement” in the online classroom, we’re creating a playbook for AI. If “engagement” or “feedback” are only a pattern of actions—a cluster of trackable things—AI can learn it and do it.
So, what can we do to “futureproof” our contribution to the classroom, the very essence of our teaching?
Finding the “handprints”
We need to get better at doing the kinds of uniquely human things that only we can do. We need to lean into our voice, our experience, our compassion, and forge real social and emotional engagement. And, again, I’m not talking about the kind of engagement that can be tracked through clicks and open rates, but engagement that blossoms in the midst of truly meaningful interactions and exchanges.
Of course, there are a lot of ways to add high-touch engagement and personalized feedback to our teaching—you can probably think of dozens. As Dr. Lo says, “Education is a field inherently linked to the human touch—one that thrives on interpersonal connections, emotional intelligence and a deep understanding of the needs of each individual learner.”
The human touch should be everywhere in our classes. But as my experiment with ChatGPT shows, some of the things we think of as our unique intellectual elaboration (like discussion board responses or assignment feedback) can be faked.
I challenge you to perform an audit on your own most recent online classroom. Try to locate that indelible stamp that only you can make. Not just the things you did do, but the things that only you could have done. Roose calls them “handprints,” those locations and actions that are uniquely human because they draw upon skill, creativity, knowledge, and emotion.
I started listing things I have done in one of my recent courses but had to repeatedly strike them through. It was a humbling process.
Attention and attentiveness
Ultimately, I decided to wipe the slate clean. When my next class started, I interrogated everything I did—every action I took, every email I sent, every ounce of intellectual elaboration. And instead of thinking about whether I could do something more or better, I started to wonder whether I could be more human in how I did everything I did.
As one example, I recorded myself grading my student’s papers, and then I sent those videos to my students, alongside more traditional feedback.
The technical and logistical considerations to enacting this were significant, and are probably different for all of us as we use different platforms, delivery methods, and different operating systems. I won’t go into it all here because the restrictions and constrictions that pop up as you develop a new idea are as much a part it as the initial spark. I will say I kind of enjoyed it, and I’m happy to share my journey (feel free to drop me an email at firstname.lastname@example.org).
The point is that students saw my face and heard my voice as I interacted with their work. They got to see how I reacted to sentences they wrote and the research they conducted. And they got to listen as I pointed out mistakes, triumphs, and opportunities to improve and grow; they could understand not just the end result but the flow of the mind at work.
Even if “encouragement” can be programmed, my voice, enthusiasm, expression, and expressiveness can’t be. I also tried to streamline my feedback in these videos. Comments and rubrics were embedded within the paper, but audibly, I focused on articulating one specific thing the student did well and one specific thing they needed to work on.
I share this with you because I want you to do everything you can to make your human-self irreplaceable, but I also want to challenge you! Take a look at your classroom and see if you can identify where you are in it. As Roose says, “The most valuable skills and abilities [are] the ones that [can] distinguish workers from machines.” We can never be faster or more comprehensive than AI.
But it is our responsibility to be unabashedly, unreservedly, conspicuously human.
Dr. Nathan Pritts is an award-winning educator, course developer, and faculty mentor with a strong focus on innovation with practical applications. He brings expertise in writing, business communication, advertising & marketing, and online user experience to the general education classroom to maximize student learning and heighten engagement, infusing curriculum with foundational outcomes bolstered by clear ties to a student’s academic and career path. Professor and program chair of First-Year Writing at the University of Arizona Global Campus, Dr. Pritts is also the author or co-author of twelve books including Decoherence (Indiana University Press), Film: From Watching to Seeing (3e), Research and Writing (2e), and Essentials of Academic Writing (4e). He also served as editor, and wrote the introduction for, Living Online: A Digital Fluency Handbook.