2023 LC Thread - It was predetermined that I would change the thread title (Part 1)

Let’s say I’m a high school student and I need to write a paper for English class? What kind of grade can I expect if I get the AI to do it? Chances of being busted for plagiarism?

What about for a college student?

1 Like

This is already happening. I assume you would get good grades, especially if you did a little bit of editing.

What about plagiarism checks?

Does the AI do a lot of copy and pasting (like a student might do) or does it actually write original content?

this sounds like great fun, except for the fact that if i even think about nextdoor, i get 4 emails asking where i’ve been and that my neighbors have been having conversations without me…

plagiarism filters also already exist, so it’s going to become an arms race between algorithms pretty quickly

https://twitter.com/itstimconnors/status/1599544717943123969

9 Likes

huh weird it didn’t claim even once that the other person just cannot read at a fifth grade level of comprehension.

2 Likes

Still waiting for the police to catch whoever killed those Idaho college students. It’s so creepy that whoever did it killed some people in the same house but not others and it seems like the cops have no leads

1 Like

If the AI generates content “in it’s own words”, then I don’t see how plagiarism algorigthms would detect it. It also wouldn’t really be plagiarism. It would be like you paid someone else to write a paper for you, which is also cheating but different from plagiarism.

This is the truth, although from what I’ve seen, the plagiarism filters are never going to be able to keep up.

Even this simple paraphraser (which is kinda cool cuz it sorta shows you how it does it in real time) is probably ahead of most of the filters.

the next generation of plagiarism filters will have to catch paragraphs that could have been generated by chatGPT from original material, and so on. in part, i am sure the models have some random choices in what they do, never try to have quite the same wording every time, but we basically still think of them as semi-deterministic black-boxes that has a a finite number of forms it can produce.

but it is totally conceivable that a genuine person writing their own thesis might use AI to complete paragraphs and then proof/edit the text, whereas a un-scrupulous plagiarizer would pull sentences of an original and simply feed them into a rewriter. they are both using a tool on OpenAI, so, the plagiarism consists simply on whether they are feeding input into it that has already been published. it can very quickly become a philosophical discussion on what is the stealing of ideas, and i don’t want to go there. i’m just hypothetically thinking that in the world of deep-fakes created by AI, we will also need AI algorithms that can identify fakes because it will be beyond what a human can analyse.

Sometimes I feel like the articles I edit are written by bots choosing random words pieced together randomly.

1 Like

https://twitter.com/TreborRhurbarb/status/1599688179455774720

2 Likes

Lol.

EubD

17 Likes
5 Likes

Isn’t there a UP discord?

Is this in the cards? I immediately mute all drama threads.

1 Like

I see historical references to at least 2 Discord servers (not counting the active crypto discord), both of which were created by @wirelessgrinder. Both invites are now invalid, presumably since they were created so long ago. I know I was in at least one of them but it was very dead and I eventually must have closed it.

There’s also an unstuck twitter account (seems inactive for the past year?) that could presumably be used for info if the webserver goes down. https://twitter.com/unstuckpolitics

I don’t think anything’s imminent but there is some chatter related to some possible legal action amidst the ongoing drama of the past couple of days. :man_shrugging:

I don’t read any of them but it appears that the mod threads are getting like 95% of the new posts so probably :harold:

1 Like