Why Banning ChatGPT in Class Is a Mistake

• 01/18/23

So, what does embracing ChatGPT look like? While I'm excited to think about that, lets start with what it doesn't look like. It's not banning ChatGPT, firewalling it, or forcing students to write in short, finite amounts of time in the artificial and distracting environment of a classroom. What it doesn't look like is a commitment to obsolescence and a refusal to accept what's only years away. A few short years from now, we (people, humans) will not be doing most of the writing of society. News stories, scientific papers, opinion pieces, fiction … most of it will be written by AI. Weird, right? But, true. I like to write and so, selfishly, I'm not too happy about ChatGPT and all it portends. But the future is here, like it or not, and AI just became a much better writer than most of us.

If that's what embracing ChatGPT doesn't look like, then what is embracing it? For starters, it's an acknowledgement of just that: "the future is here, like it or not, and AI just became a much better writer than most of us." Teaching our students to write is just like teaching our students arithmetic without a calculator or literature research without the internet. Should we teach that? Yes … once or twice. And then coach the leveraging of technology. As I sit here clacking away at my laptop keyboard, I think of my editor who will soon be reading this piece. Five years from now, she'll have her job and I won't have mine (at least not this job). The content that AI generates will always need to be edited, sourced, fact-checked, and massaged into the appropriate voice, theme, and message for the target audience. Humans are good at intuition, feel, and subtleties (in this context: editing), and that's what AI will lack for a long time still.

So let's model that, coach it and teach it to our students — the editing and analyzing of AI-generated writing. Instead of banning AI in the college classroom, I'm planning to create assignments where students must prompt ChatGPT to draft a research paper, opinion piece, or reflection, and then the student must source that paper and fact-check it as the human component of the assignment. Another idea that I'm toying with is breaking students into groups and assigning each opposing perspectives of controversial topics. Those groups would then use ChatGPT to draft talking points related to their assigned positions, and use those talking points as starting points for real time, in-class discussions. A third idea: Have students write a short paper on an assigned topic and use ChatGPT to do the same. Students then exchange these paper-pairs (one human-written and the other AI) and do a peer-review of both papers in the pair, trying to identify which is ChatGPT and why they think so. The common theme in all of these assignment ideas is the analysis of the product created by ChatGPT. Thinking critically about the content that AI produces will be essential for the next generation, and we should be teaching it today. Educators avoid this path at the peril of our own obsolescence; something higher education is already at far too much risk of already.

"In conclusion, ChatGPT has the potential to revolutionize the way college classrooms operate by providing a unique learning platform for students to engage with and ask questions. With its ability to generate human-like responses to student inquiries, ChatGPT can help facilitate discussions and provide personalized learning experiences for students. As more educational institutions begin to adopt ChatGPT and other AI technologies, it will be interesting to see how these tools continue to shape the future of higher education." (Oh, and by the way, I didn't write that last paragraph. ChatGPT did with the prompt: "write a closing paragraph for a piece on the role of ChatGPT in the college classroom.")