[thinkingemote]

Recently there was a flood of articles about "students are using AI to cheat". Now there's a flood of articles about "students are anti AI".

My first impression is that floods of articles do not accurately reflect the real world, but just show some facet of it. But if they are both correct and both are to be taken as real, should we expect that students will agree with academia and not use AI in their education? Might we see the return of traditional learning?

(Education is different than our industry. In our industry, most of those using LLMs are forced to by the powers to be. In education, the powers to be do not want the students to use LLMs.)

[kartoffelsaft]

As a recent graduate, both headlines read as plausible and non-contradictory. Students right now are faced with two things:

1. a machine that can do the things asked of them faster, more accurately, and higher quality.

2. the threat that that machine completely or mostly invalidates their education, in particular for getting an entry-level job because they don't exist anymore.

The former headline is a result of point 1 and the latter point 2. They're using it not because they think "it's good, actually" but because they're resigning themselves to their education not mattering for their professional development and taking the easy path. That breeds the resentment that you see with "students are anti-AI".

[827a]

One of the "grown up" moments everyone needs to make their way through is: Realizing that the vast majority of people are not internally consistent, and that by the way includes you.

Every single student who boo'd Eric Schmidt the other day was regularly using AI for their schoolwork. People are not cistercian monks.

Its easy to draw a conclusion from this like "revealed preferences outweigh spoken ones, we can ignore the boos" but much like the tech executives, you're not thinking deeply enough. The tech industry will face the music for relentlessly creating products that the world hates to be forced to use. But, for now, the industry is too addicted to it. It sounds crazy, but: There are vanishingly few companies left who have the ability to manufacture products & services that their customers are excited to use. Its a lot easier to monopolize a space, re-baseline the industry around the expectations of your product's existence, then deploy capital and lawyers to put up fences.

[InsideOutSanta]

If some students use LLMs to do tasks faster and at higher quality, that changes the grading curve, so everybody else might have no choice but to do so as well if they want to graduate. It's the "and yet you participate in society" meme.

[limflick]

I don't think those two things are mutually exclusive. Good chance that a few students that cheated or at the least used AI in a major capacity to graduate, still booed when that former Google CEO brought up AI at the graduation speech. Being pro AI when it benefits them and anti AI when it doesn't is just human nature. I'm being a little reductive here though.

[mold_aid]

Both happen at the same time, by the same people. The reason for usage differs by major, but usually it's an expedient to either get past tasks that represent busywork or just the cheating you've seen described. Students have explained to me how much they hate in the same explanation of what they do with it.

(just FYI: There's no "traditional learning" to return to; you will definitely hear a lot of faculty going to "paper and pen" situations - kinda uncritically, if you ask me! - but I ask folks to remember that writing itself is a technology, and the media/means historically associated with it are technological advances in their own rights).

[HarHarVeryFunny]

> Recently there was a flood of articles about "students are using AI to cheat". Now there's a flood of articles about "students are anti AI".

Sure, both are true, although I think you'll find that they differentiate between "cheating" on their math homework by using AI, and kids who are cheating on exams by sneaking in a smartphone and giving a photo of the problem to ChatGPT.

As far as homework goes, AI is just the new Google, useful perhaps, but hardly outweighing all the anxiety of their future being taken away by AI, or all the societal enshittification by AI that they see all around them.