Excessive use of words like ‘commendable’ and ‘meticulous’ suggests ChatGPT has been used in thousands of scientific studies
https://english.elpais.com/science-tech/2024-04-25/excessive-use-of-words-like-commendable-and-meticulous-suggest-chatgpt-has-been-used-in-thousands-of-scientific-studies.html
120 points · 16 comments · view on lemmy.world
16 Comments
aBundleOfFerrets@sh.itjust.works · 58 pts · 2y
I like how the dick rat has become a beacon of scientific malpractice.
hydroptic@sopuli.xyz · 30 pts · 2y
dohpaz42@lemmy.world · 23 pts · 2y
I’m using ChatGPT to help me simplify the very terse language of an academic paper, and I must say I’m super unimpressed so far. I don’t understand how people could possibly use it to write anything of substance given the output it generates; it generates redundant, overlapping, and superficial responses that need to be heavily edited to make sense. I’m pretty much better off trying to decipher the paper by myself.
BirdEnjoyer@kbin.social · 5 pts · 2y
canihasaccount@lemmy.world · 2 pts · 2y
It can't write much of substance. The only people using it in science for anything more than fluff are people who don't speak English well or who have no business writing papers. I sympathize with the former, but I don't understand why those folks wouldn't just either publish in a language they speak or get an English-speaking coauthor to help write in English. I wouldn't ever use it to write an article. Even editing, it tends to butcher scientific nuance.
It is good at writing fluff though, which is helpful for things like letters of recommendation for undergraduates.
hydroptic@sopuli.xyz · 1 pts · 2y
canihasaccount@lemmy.world · 1 pts · 2y
I'm in science. It isn't difficult to get an English speaking coauthor. Going to an LLM is easier and faster, sure, but if someone can't understand the output then they have no idea if their text is being translated correctly.
hydroptic@sopuli.xyz · 0 pts · 2y
canihasaccount@lemmy.world · 1 pts · 2y
Ah, I would consider that fluff, which is okay in my book. I don't use it for writing, personally, but what I tell my students is that if it'd be fine for a friend to do the thing and not get coauthorship, it's fine to use AI for that (provided you acknowledge it, as you would a friend who provides some helpful comments on a draft). Proofing and suggesting minor stylistic things fall under that umbrella IMO.
hydroptic@sopuli.xyz · 1 pts · 2y
Tar_alcaran@sh.itjust.works · 1 pts · 2y
LLMs can only copy information, they don't evaluate it. So they end up with a quality level around the median of the input. And since most content is pretty crappy, you end up with mediocre crappy output.
Thankfully, there's very high demand for mediocre crap (which is why there's so much of it).
AnUnusualRelic@lemmy.world · 1 pts · 2y
Well, of course if you want substance...
Semi_Hemi_Demigod@lemmy.world · 18 pts · 2y
ChatGPT is now at the intelligence of a sleep deprived undergrad, using the word "terrible" five times in one run-on sentence the morning before the paper is due
nucleative@lemmy.world · 9 pts · 2y
Soon, ChatGPT will train itself on these articles leading to it gaining a stronger and stronger probability of using words like 'meticulous' and 'commendable'. Soon all output will be non stop repetition of 'commendable'.
esc27@lemmy.world · 5 pts · 2y
I'm wondering if this is happening with people as well. If some of these could be written by humans who subconsciously picked up AI phrasing by reading to much AI text.
Aradina@lemmy.ml · 2 pts · 2y
Commendaticulous you mean
venoft@lemmy.world · 8 pts · 2y
Using chatgpt is not bad in itself, but then you have to make sure to meticulously read the generated article to see if it's all correct. It's commendable people want to make sure it's easy to read and not mostly changrish or so. An article is just for the transfer of knowledge anyway, who or what writes it doesn't matter as long as the facts are correct.
A_A@lemmy.world · 0 pts · 2y
Using chat GPT doesn't change chat GPT.A language model system learn during a phase that's called "training". After the training the system doesn't learn. We, as users, interact with it only after training.
Edit : Oops ! My comment above was out of context because I didn't read the article ... only the title ... and I misinterpreted it as meaning that "
chat GPT overused certain words" which is not what the article says.empireOfLove2@lemmy.dbzer0.com · 9 pts · 2y
You think OpenAI has just completely stopped trawling the internet for content? Nobody knows how often they update chatGPT, it's likely a lot.
A_A@lemmy.world · 1 pts · 2y
Oops, my bad ... so I corrected my above comment.
blargerer@kbin.social · 3 pts · 2y
A small nuanced correction to your comment about language models being pre-trained. That's true for GPT (its the P), and all modern high performance language models since 2017, but it was actually not in fashion before that period, and may not be again at some point in the future.