Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

news.fsu.edu

Post reply on HN
Print view View original

51–60 of 381 posts

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#51

Post by rogerrogerr »

Aurornis wrote: post #37
rogerrogerr wrote: post #33

Earlier quoted context omitted.

You should be thankful for the AI “accusers”; most of us will just assume you used the slop machine and stop reading whatever you wrote without wasting our breath telling you about it.

> most of us will just assume you used the slop machine and stop reading whatever you wrote From what I've seen, the people who jump to hasty conclusions about AI use mostly do it when they disagree with the content. When the writing matches what they want to see, their AI detector sensitivity goes way down.

Wouldn't surprise me if that's true. I just treat any AI-smelling content as an information hazard that is _at best_ providing no useful entropy and stop reading it. Something about it is just so repulsive.

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#52

Post by QuantumNomad_ »

dakiol wrote: post #14
mrbonner wrote: post #7

I intentionally put spelling mistakes in my doc to let others know I'm not using ChatGPT. What a time to be alive in which small spelling or grammar mistake is a good sign of authenticity.

Same. Also, when asked for anonymity at work, I usually make mistakes that do not correspond to my native tongue (let’s say I’m french and working in an international company. I would write comments in a supposedly anonymous survey like “He ist like…” to camouflage myself as german). It’s so easy to trick everyone. People who doesn’t do that is just too lazy. In slack, you cannot just copy paste a two-paragraph answe…

> I would write comments in a supposedly anonymous survey like “He ist like…” to camouflage myself as german

Do actual Germans ever make that kind of mistake though?

I’ve only ever seen “ist” used “wrongly” in that particular way by English speakers, for example in a blog post title that they want to remain completely legible to other English speakers while also trying to make it look like something German as a reference or a joke.

The only situation I could imagine where a German would accidentally put “ist” instead of “is”, is if they were typing on their phone and accidentally or unknowingly had language set to German and their phone autocorrected it.

Sometimes you get weird small things like that on some phones where the phone has “learned” to add most English words to the dictionary or is trying to intelligently recognise that the language being written is not matching the chosen language, but it still autocorrects some words to something else from the chosen language.

But I assume that when people fill out forms for work, they are typing on the work computer and not from their phone.

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#53

Post by oasisaimlessly »

Aurornis wrote: post #37
rogerrogerr wrote: post #33

Earlier quoted context omitted.

You should be thankful for the AI “accusers”; most of us will just assume you used the slop machine and stop reading whatever you wrote without wasting our breath telling you about it.

> most of us will just assume you used the slop machine and stop reading whatever you wrote From what I've seen, the people who jump to hasty conclusions about AI use mostly do it when they disagree with the content. When the writing matches what they want to see, their AI detector sensitivity goes way down.

Yeah, that's human nature.

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#54

Post by ASalazarMX »

dgfitz wrote: post #41
jazzypants wrote: post #16

Earlier quoted context omitted.

"The dwarves delved too greedily and too deep." - Saruman, 2002

"The Dwarves tell no tale; but even as mithril was the foundation of their wealth, so also it was their destruction: they delved too greedily and too deep, and disturbed that from which they fled, Durin's Bane" - J.R.R. Tolkien spoken by Gandalf, 1954

scoff It's evident that Gandalf clearly used AI. Saruman is the real human here.

Jokes aside, I don't like what LLMs are doing to our culture, but I'm curious about the future.

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#55

Post by Taek »

diego_sandoval wrote: post #10
milancurcic wrote: post #4

"Recent large-scale upticks in the use of words like “delve” and “intricate” in certain fields, especially education and academic writing, are attributed to the widespread introduction of LLMs with a chat function, like ChatGPT, that overuses those buzzwords." OK, but please don't do what pg did a year or so ago and dismiss anyone who wrote "delve" as AI writing. I've been using "delve" in speech for 15+ years. It's…

Same thing as with em dashes. Some of us have been using em dashes from before ChatGPT.

Genuine question, do you actually use the formal emdash in your writing? AIs are very consistent about using the proper emdash—a double long dash with no spaces around it, whereas humans almost always tend to use a slang version - a single dash with spaces around it. That's because most keyboards don't have an emdash key, and few people even know how to produce an actual emdash.

That's what makes it such a good giveaway. I'm happy to be told that I'm wrong, and that you do actually use the proper double long dash in your writing, but I'm guessing that you actually use the human slang for an emdash, which is visually different and easily sets your writing apart as not AI writing!

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#56

Post by dragonwriter »

jgalt212 wrote: post #12
milancurcic wrote: post #4

"Recent large-scale upticks in the use of words like “delve” and “intricate” in certain fields, especially education and academic writing, are attributed to the widespread introduction of LLMs with a chat function, like ChatGPT, that overuses those buzzwords." OK, but please don't do what pg did a year or so ago and dismiss anyone who wrote "delve" as AI writing. I've been using "delve" in speech for 15+ years. It's…

Fair enough, but if you know you're audience may be dismissive of your writing and its message if you use such words, it behooves one to steer clear of AI slop words. IIRC, such offenses in school writing are tagged PWC (poor word choice).

The thing is virtually every single thing that gets presented as an "AI tell" is just "a word, punctuation mark, or pattern of presenting information more common in a training set which includes a high volume of formal writing and professional presentations than it is in the experience of people whose reading and writing is mostly limited to social media and low-effort listicle-level online 'journalism'."

So, yeah, if your target audience are the people who take those "AI tells" seriously and negatively react to them, definitely craft your writing to that audience. But also, consider if that is really your target audience...

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#57

Post by ACCount37 »

freehorse wrote: post #15

I had first noticed "meticulous" to be used a lot in translations from chinese. Is it sth about chinese itself (that they use sth a lot for which meticulous is the closest translation), or about some translation software that is possibly biased towards such buzzwords when translating to english?

A lot of those "ESL" patterns are cultural.

It's a mix of a cultural "founder effect" - whoever writes the English textbooks and the dictionaries gets to shape how English is learned in a given country - and also the usage patterns of the source language seeping through. In your case, it's mostly the latter.

Chinese has a common word with a fairly broad meaning, which often gets translated as "meticulous". Both by inexperienced humans and by translation software.

Ironically, a few Chinese LLMs replicate those Chinese patterns when speaking English. They had enough "clean" English in their pre-training datasets to be able to speak English. But LLMs are SFT'd with human-picked "golden" samples and trained with RLHF - using feedback from human evaluators. So Chinese evaluators probably shifted the LLMs towards "English with Chinese ESL influence".

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#58

Post by dotinvoke »

Taek wrote: post #55
diego_sandoval wrote: post #10

Earlier quoted context omitted.

Same thing as with em dashes. Some of us have been using em dashes from before ChatGPT.

Genuine question, do you actually use the formal emdash in your writing? AIs are very consistent about using the proper emdash—a double long dash with no spaces around it, whereas humans almost always tend to use a slang version - a single dash with spaces around it. That's because most keyboards don't have an emdash key, and few people even know how to produce an actual emdash. That's what makes it such a good givea…

A mobile keyboard—limited as it is—has no trouble producing an em-dash, requiring little more than a long press on the - button.

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#59

Post by ASalazarMX »

yesco wrote: post #19
mrbonner wrote: post #7

I intentionally put spelling mistakes in my doc to let others know I'm not using ChatGPT. What a time to be alive in which small spelling or grammar mistake is a good sign of authenticity.

I understand people being paranoid about this, but just understand that the people who will judge you for spelling errors will always dwarf the ones who believe they are capable of sniffing LLMs out...

Besides, it's too easy to ask LLMs to add a few spelling and grammar mistakes.

Re: Researchers find evidence of ChatGPT buzzwords turning up in everyday speech

#60

Post by adastra22 »

Fade_Dance wrote: post #39
diego_sandoval wrote: post #10

Earlier quoted context omitted.

Same thing as with em dashes. Some of us have been using em dashes from before ChatGPT.

Unfortunately the em dash has already been relegated to the dungeon of AI suspicion for the next 5-10 years.

I often edit things in Word — I have a document that I can alt-tab to and type things. It has spellcheck, etc. that my browser window does not, and I’m not at risk of losing if I refresh or something. Then copy-paste back.

Word converts any - into an em dash based on context. Guess who’s always accused of being a bot?

The thing is, AI learned to use these things because it is good typographical style represented in its training set.

Post reply on HN