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

311–320 of 381 posts

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

#311

Post by lo_zamoyski »

ACCount37 wrote: post #114
bonoboTP wrote: post #102

Earlier quoted context omitted.

> In the "opinion" of ChatGPT, my style of writing is "academic". It may simply be glazing. If you ask it to estimate your IQ (if it complies), it will likely say >130 regardless of what you actually wrote. RLHF taught it that users like being praised.

And, if you want to have some fun, you could give it your writing sample - but say that it's from a random blog post you found online. See what it tells you on that. It really is a shame that an average user loves being glazed so much. Professional RLHF evaluators are a bit better about this kind of thing, but the moment you begin to funnel in-the-wild thumbs-up/thumbs-down feedback from the real users into your trai…

Not everyone appreciates having his speech characterized as "academic" - in certain circles, it's viewed rather poorly - so I'm not convinced of the glazing hypothesis.

ChatGPT certainly makes distinctions. If I give it a blog post written by a philosophy professor, I get "formal, academic, and analytical". If I feed it an article from The Register, I get "informal and conversational". The justifications it gives are accurate.

"Academic" may simply mean that your writing is best characterized as an example of clearly written prose with an expository flavor, and devoid of regional and working class slang as well as any colloquialisms. Which, again, points to my RP comparison.

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

#312

Post by Angostura »

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.

The article also uses M dashes

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

#314

Post by ModernMech »

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'm writing fully conversationally now, in my own voice. If people want a particular tone, they can ask ChatGPT for one.

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

#315

Post by the_af »

viccis wrote: post #148
Fade_Dance wrote: post #39

Earlier quoted context omitted.

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

Good. It's a crutch for poorly composed sentences or for prose intending to imitate the affect of poorly composed sentences. There's not a single sentence under the sun that needs an emdash. Commas and parentheses can do it all, and an excess of either is a sign of poorly edited prose. I don't buy the pro-clanker pro-em dash movement that has come out of nowhere in the past several years.

> There's not a single sentence under the sun that needs an emdash

Sentences "need" very little, but without style and personality, writing becomes very boring. I suppose simplicity without any affectation works for raw communication of plain technical facts, but there's more to writing than that.

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

#316

Post by redwall_hp »

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…

Option + shift + hyphen or hold hyphen on any Mac or iPhone to get an em dash. I use them very frequently, because they're the correct character for the use case.

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

#317

Post by staticman2 »

lo_zamoyski wrote: post #308
bonoboTP wrote: post #102

Earlier quoted context omitted.

> In the "opinion" of ChatGPT, my style of writing is "academic". It may simply be glazing. If you ask it to estimate your IQ (if it complies), it will likely say >130 regardless of what you actually wrote. RLHF taught it that users like being praised.

That assumes the characterization is perceived as flattering, or that enough data on me would allow it to "think" it would be to me. Generally, given the anti-intellectual bias in American popular culture, I'm on the fence about that. But then, what are the biases of the corpus ChatGPT was trained on? For context, I was asking GPT to rewrite some passage in the style of various authors, like Hemingway or Waugh. I did…

Even if, for some reason, you think LLM's are fit for evaluating writing style (I don't), I'd at least ask Gemini Pro and Claude Opus to see if there's consensus among the plausible sounding bullshit generators.

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

#318

Post by ACCount37 »

lo_zamoyski wrote: post #311
ACCount37 wrote: post #114

Earlier quoted context omitted.

And, if you want to have some fun, you could give it your writing sample - but say that it's from a random blog post you found online. See what it tells you on that. It really is a shame that an average user loves being glazed so much. Professional RLHF evaluators are a bit better about this kind of thing, but the moment you begin to funnel in-the-wild thumbs-up/thumbs-down feedback from the real users into your trai…

Not everyone appreciates having his speech characterized as "academic" - in certain circles, it's viewed rather poorly - so I'm not convinced of the glazing hypothesis. ChatGPT certainly makes distinctions. If I give it a blog post written by a philosophy professor, I get "formal, academic, and analytical". If I feed it an article from The Register, I get "informal and conversational". The justifications it gives are…

Does an average user appreciate this?

Do you?

The first question matters because frying an AI with RL on user feedback means that the preferences of an average user matter a lot to it.

The second question matters because any LLM is incredibly good at squeezing all the little bits of information out of context data. And the data you just gave it was a sample of your writing. Give enough data like that to a sufficiently capable AI and it'll see into your soul.

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

#319

Post by halper »

DrBazza wrote: post #275

It could be worse, it could be 'learnings'. It's lessons. We don't go for 'drivings'. Though ChatGPT will probably force more nonsense like that into the mainstream. That's from the last decade. 'Please revert' seems to be from the 00's, it's 'reply'. There are others I've tried to ignore and forget. Language changes, and I'm a dinosaur unfortunately. I also love the fact American English sometimes uses better, or mo…

I have made a learning and it is that one could have an ask.

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

#320

Post by Procrastes »

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…

I have --- set to autocorrect to —. I've been using it in formal writing for 30 years. When we were in high school, we had a "Dash Party" in English class, where we ate Twinkies and learned about the different dashes.

I would argue that LLMs overuse the emdash more because they overuse specific rhetorical devices, e.g. antithesis, than because they are being too correct about punctuation.

Post reply on HN