> > a skilled LLM user can generate solid responses to almost anything.

> I don't think my students, who often leave the "as an AI I have no opinions" in their essays, are skilled LLM users

This thread is really funny; or really depressing, depending on your point of view. My conclusion: teachers should start using ChatGPT to grade these assignements. Without even reading them. That would be a win-win for everybody.

My takeaway from the thread is that LLM-use seems to be the manifestation of student's disinterest, dissatisfaction, and potentially poor attention spans. Comment from one student (who, from what I glean from the thread, is probably more capable than average).

There is one comment from a student in the thread. Sure they seem angry, but I also see their point:

> I have used chatgpt in literally all of my essays this semester, and got an average grade of 8-9.

> ...

> But you know why, at least in my case, I always get away with it? Because all of the work I have to do requires ZERO critical thinking from my side. In fact, it is even worse to use it sometimes. Don't you even dare to come up with your own opinions!!! Just quote someone that is already quoting someone and so and so on.

> I know most students don't give a single flick about your class and that unmotivates but damn it is the 5th year you ask for the same essay and we know you don't even read half of them. Why would I lose so much time doing something literally no one cares about.

> TL;DR: If a machine can do most college essays it's because they were mechanical jobs already.

It's strange, all these professors asking students to generate outlines, when outlines are probably the worst way to write an essay. The kids don't want to write because they have no investment in it: professors ask for systematic, boring, easily interpretable essays, they get essays generated by a program that writes in the most easily interpretable way possible.