ChatGPT and the value of authenticity

This last semester, I had a number of student essays that I suspected were written by ChatGPT or some other AI text machine, and one that was confirmed as such — it included hallucinated content. It’s hard to prove the use of ChatGPT or similar, so I let it go, except for the instance with fictive facts. AI checkers are notoriously spotty, and I don’t want to police student writing, and I definitely don’t want to accuse students of cheating on the basis of false positive AI checker results.

I am not satisfied by this state of affairs. I don’t mind if AI machines are used as tools, for example to correct student writing, or to provide references on a topic, or such. What worries me (and what let me to suspect AI use) is the production of generic but correct, bland text. Students don’t do the work that (I am convinced) would help them learn something, they forgo the opportunity to think about a topic, and they produce generic ticker-tape of an essay that is easy to read but that nobody really wants to read.

I believe that one of the tasks for us in this age of AI is to find ways for us humans to be authentic, and to be so in a way that cannot be imitated by a machine. For students, this means to be authentically imperfect and to accept those imperfections, until they have found their own voice that allows them to be authentically perfect if they choose so. (I don’t know if I want my writing to be perfect, authentically or not.) I suspect there has to be some punk in our writing.