What is the meta question about ChatGPT?

Meta (from the Greek μετά, meta, meaning "after" or "beyond") is a prefix meaning "more comprehensive" or "transcending". (Wikipedia)

Predictably, amid the hand-wringing and angst in academic circles whenever the topic of ChatGPT comes up, someone is designing ways to police it's use and detect AI-assisted cheating.  Fine, and perhaps necessary, given that there is much to figure out about how AI will change teaching and learning. But I am reminded about the hand-wringing and angst among teachers and parents when Texas Instruments and Hewlett-Packard introduced relatively inexpensive mass market programmable hand calculators.

Simply banning using ChatGPT or developing ways to detect its use is a classic "single loop" (or knee jerk) reaction to something that begs more of a meta analysis. The field needs to quickly  move from, "how do I find out if my students are using ChatPGT on their AP history paper," to a higher-order question about what AI means for how we assess learning. Indeed, maybe the meta question is really about what "learning" means in an AI-assisted environment.

There was a time when hand calculating long division in mathematics or computing a sum of squares in statistics constituted an assessment of learning. Programmable calculators and statistical packages such as SPSS or SAS took us past those days. What will AI take us past?