ChatGPT, Take Two

[andrea_lunsford]

A few months ago, I wrote in this space about the rise of AI chatbots like ChatGPT. Since then—like you, I imagine—I have been inundated with articles and YouTube videos and tweets about the still fairly new Open AI chatbot. We are seeing an understandable amount of handwringing among the academic community as administrators and teachers everywhere worry about academic integrity, about “the end of writing,” and even about their jobs. As with every sudden technological innovation, this one is unsettling at best and perhaps terrifying at worst. I have read a lot of the jeremiads and the end-of-the-world-as-we-have-known-it scenarios, and while I am sympathetic, I am at least for the time being inclined less to a “the sky is falling” stance and more to a cautious, careful look at how we and our students can work productively, and ethically, with this new tool. If the slide rule and the calculator didn’t bring the world to a halt, if television did not destroy education, then it seems unlikely that ChatGPT will do so.

A recent New York Times piece takes this position. Author Kevin Roose recognizes legitimate fears about plagiarism and other ethical issues but goes on argue that “schools should thoughtfully embrace ChatGPT as a teaching aid—one that could unlock student creativity, offer personalized tutoring, and better prepare students to work alongside A.I. systems as adults.” He makes this argument for three main reasons: 1) that ignoring or trying to keep such AI tools away from students simply will not work: they are here and here to stay; 2) that ChatGPT can be an “effective teaching tool,” giving the example of a teacher whose students use the chatbot to create outlines for essays they are working on—and then go on to write stronger and more engaging essays as a result, or using the tool as an out of class tutor or a debate sparring partner or “starting point for in-class exercises"; and 3) that ChatGPT helps teach students about the world they live in, one full to bursting with AI technologies they will need to “know their way around . . . their strengths and weaknesses, their hallmarks and blind spots—in order to work alongside them.”

I found Roose’s argument thoughtful and even-handed and his examples compelling. Still for me, in these fairly early days, I find that ChatGPT and tools like it seem most perfectly suited for play, for the ludic possibilities of rhetoric—as the example of students using ChatGPT to compose satiric love notes and poems demonstrates. But play, remember, is not “mere” play but a crucial part of learning and cognitive development. “Playing around” with language and other symbols is fundamental to growth. So my inclination is to encourage students to play with this new tool, pushing it to its limits, using it to entertain and amuse, all the while using such activities to learn about its strengths and weaknesses, the “blind spots” Roose mentions, and always testing it against their own creativity, ingenuity, and knowledge. We teachers of writing will have a lot to learn from such play!

Image by Jonathan Kemper reproduced under the Unsplash license.