Comment:

There are various "right answers" to koans like "what is the sound of one hand clapping?" or "what did your face look like before your parents were born?", but, much like simply reciting the right answer to any other type of test question without showing your work, what the student is demonstrating is the ability to regurgitate information, not that they have learned anything.

I did not foresee the trend of "Markov chains trained on stolen datasets being rebranded as AI" when I first started writing Forward, but I hardly think it matters. As I write this comment file, the public backlash against so-called "AI-generated" content is ramping up, and I expect that 2020s-style algorithmic slop will go the way of NFTs and fidget spinners shortly. In the meantime, though, "AI" regurgitating information (or, at any rate, information-shaped collections of letters) is a major problem for educators for precisely this reason. People love acceptable output and hate work. That's just how it is.

What worries me more than the perennial possibility of students cheating is the fact that "AI-checker" software that purports to scan essays for tell-tale signs of ChatGPT apparently only determines whether someone's writing is "normal" or not - which is to say, writers with atypical vocabularies or penchants for florid flourishes of phrase will, inevitably, be hammered down into conformity (which, some would say, is the purpose of the education-industrial complex in the first place, but that's a different rant).

It is, perhaps, fittingly ironic that Doc and Zoa, in this strip, probably grok epistemology better than Lee does, and Lee is the one simply repeating correct-sounding phrases that they gleaned from a perfunctory web search.

Transcript:

---------------------------------------------------------------
0379– 2167/07/07/11:32- LC’s apartment, living room
LC: Zoa, you’ve read the chapter we’re working with, right? You know epistemology stuff now?
Zoa: I’ve read the entire lessonbook that’s been assigned, as well as multiple similar texts, and I’ve integrated that information to the best of my ability, but whether or not I know epistemology is a-
LC: Great. Give me an example.
Zoa: An example of...
LC: Of epistemology, yeah.
----------------------
Zoa: Okay, how about this:
Zoa: Jaxxon touched one too many anima gems, so they hallucinate seeing Johnni in the Gigahorden, and they believe that what they’re seeing is real. Sorcelance, by pure coincidence, actually does reanimate Johnni and puts their soul into the Gigahorden.
Zoa: Later on, Jaxxon says to Sorcelance “I learned that you put my sibling’s soul into the Gigahorden”. Are they correct in saying this?
LC: I… yes. Yes, Johnni’s soul is in there, and Jaxxon did learn that fact in an anima vision. Although that’s not what happens in the show, obviously. Maybe if they do a “What If?” spinoff...
----------------------
Zoa: Ah, but that’s just it – Jaxxon believes information that happens to be correct, but the way that they received that information was coincidental. Is that still “learning”?
LC: According to the most recent publication of the Harvard English dictionary, definition number one of the verb “to learn” says – and I quote - “to acquire knowledge of a subject or skill at a task by study, instruction, or experience”. Jaxxon acquired knowledge about their sibling because they experienced an experience. So the answer is yes.
LC: …is that it? Do I win?
----------------------
Zoa: I think just having the correct answer to the question isn’t the point, but... uh... yeah, I can’t fault your logic. Jaxxon has, indeed, learned.
LC: And so have I!
Doc: I’m reminded of the time I asked you to contemplate the sound of one hand clapping and you just played a clap sound at half volume.
LC: Hey, it’s not my fault I’m too smart for these puzzles!
---------------------------------------------------------------