How I am Learning to Stop Worrying and Love ChatGPT

Artificial Intelligence concerns me.

As a member of the global village, I’m worried that ChatGPT, Bard, Claude, and other technologies will become ubiquitous.

As a professor, I’m more immediately concerned with how AI is impacting teaching and learning.

I first heard about ChatGPT in November 2022. After making an account and playing around with it, I forgot about it within minutes. I didn’t need a tool that could instantly write a 1000-word essay on a given prompt or answer various questions. ChatGPT didn’t seem relevant for my life.

I was naïve.

I even forgot about ChatGPT until February and March 2023 when I began to get some very unusual assignments from my undergraduate students: submissions written with an unusual tone, with odd and completely inaccurate statements, and with structures that just felt unnatural.

I quickly learned that most students knew all about ChatGPT from TikTok, including other tools designed to take what ChatGPT created and endow it with a more natural-sounding voice.

I quickly learned how profound ChatGPT is after asking it to answer some of my exam questions. Consider this (once-challenging) question from an essay exam in my Theologies from the Margins class:

What makes In My Grandmother’s House a work of womanist theology? Discuss the difference between its praxis-based approach and the somewhat more theory-based approach in Reading the Bible from the Margins. Explain how and why. Use examples, too.

A human response to such higher-order questions takes substantial time and skill—thinking, reading, synthesizing, and writing. But for ChatGPT, a response takes mere seconds. ChatGPT’s response isn’t outstanding, but it often produces a “C” or “D” level response. A student could even use ChatGPT as a shortcut by building on its response.

I have since been trying to figure out ways to respond to the challenges of AI. In conversations with colleagues, there are two basic ideas about how to react to AI.

One response is not to make any changes and that if students cheat, they are only hurting themselves.

The other general response is to make assignments ChatGPT-proof, and if students can use ChatGPT to complete an assignment, it’s a bad assignment.

Both responses concern me.

I suggest that we should always set students up for success; thus, we must care about ChatGPT.

I also offer that it is naïve to think that pedagogies can simply change overnight (or even over a semester or two) and naïve to think an assignment can be ChatGPT-proof, especially as the technology further adapts. ChatGPT will do almost anything with the correct set of commands, and there is still a place for and value in recall and other lower-order thinking skills.

If a professor is resolute in issuing an AI-proof assignment, the only option is to mandate that students complete it in a controlled environment, a high-stress environment that is antithetical to most best practices.

And it’s not just ChatGPT forcing the educational landscape to embrace a brave new world: Grammarly now incorporates AI and ChatGPT. It has gone from proofreader to co-author. (Grammarly is coming after those cherished “AI-proof assignments”!)

The following excerpts, supplied by a student to help me and others understand Grammarly’s new capabilities, come from a discussion board assignment related to Andrew L. Seidel’s excellent and thought-provoking American Crusade: How the Supreme Court Is Weaponizing Religious Freedom. While I do not agree with the student’s assessment of the book, the two paragraphs show another layer to the complexity of addressing AI when it’s interwoven with word processing programs.

Original draft:

Im glad to see that alot of us as readers noticed that the majority if not all of the case material stated as evedince in the book only supports Seidel perspective. To me that perspective being an Athiest point of view. My problem with that is , I left feeling like my choice was either all or nothing. If I support the Crusaders im clearly backing a White Supremacist pPatriachy, but if I side with Seidel I’m left feeling like Im in aggreeance with the complete absence of God at all. If the author could have been more inclusive of religious freedom with in the perameters of the lines he explained it would have created a more receptive front for a fight for a collective goal.

Draft after Grammarly:

I’m pleased to observe that many of us, as readers, have noticed the book’s reliance on case materials that predominantly support Seidel’s perspective, which can be identified as an atheist point of view. However, I have a concern regarding the binary nature of the presented choices. It appears that supporting the Crusaders is equated with endorsing a white supremacist patriarchy, while aligning with Seidel’s arguments feels like advocating for the complete absence of God. It would have been more beneficial if the author had taken a more inclusive approach to religious freedom, demonstrating that it can exist within the parameters he established. By doing so, a more receptive and unified front for addressing collective goals could have been fostered.

AI is changing how we read and write. Works such as George Orwell’s “Politics and the Use of the English Language” and Vershawn Ashanti Young’s “Should Writer’s Use They Own English?” emphasize the importance and beauty of writers using their own language such that they can understand and own their knowledge, experience, and words.

And the problem is more than pedagogical. It’s ethical. We’re facing a radical outsourcing of human cognition. Are we really willing to outsource the very things that help make us human—our abilities to use language, our abilities to think, create, and communicate? ChatGPT could well be another example of history and science being anything but a march of progress.

I get that—for students—the temptation to use ChatGPT is irresistibly tempting. It’s new. The pressures on college students of the 2020s are extraordinary: Financial strains, family responsibilities, coursework, and Long COVID (i.e., SARS-CoV-2) keep students en masse exhausted. A challenge for students will also be the various ways institutions and professors react to AI in instructional contexts.

I don’t know what will happen. I don’t have any solutions. Currently, my goal is only to capture and share these thoughts and to continue grappling with all things AI. I do know that I don’t want to read work authored by AI.