Most active commenters
  • Buttons840(3)
  • chaps(3)

←back to thread

It's rude to show AI output to people

(distantprovince.by)
321 points distantprovince | 20 comments | 19 Jul 25 16:57 UTC | HN request time: 0.001s | source | bottom
Show context
phito ◴[19 Jul 25 17:28 UTC] No.44617442[source]▶
I really wish some of my coworkers would stop using LLMs to write me emails or even Teams messages. It does feel extremely rude, to the point I don't even want to read them anymore.
replies(10): >>44617497 #>>44617500 #>>44617658 #>>44617721 #>>44617880 #>>44617940 #>>44618006 #>>44618504 #>>44619441 #>>44622817 #
benatkin ◴[19 Jul 25 17:34 UTC] No.44617500[source]▶
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
replies(1): >>44617529 #
1. echelon ◴[19 Jul 25 17:37 UTC] No.44617529[source]▶
Wow. What a good giveaway.

I wonder what others there are.

I occasionally use bullet points, emdashes (unicode, single, and double hyphens) and words like "delve". I hate it think these are the new heuristics.

I think AI is a useful tool (especially image and video models), but I've already had folks (on HN [1]!) call out my fully artisanal comments as LLM-generated. It's almost as annoying as getting low-effort LLM splurge from others.

Edit: As it turns out, cow-orkers isn't actually an LLMism. It's both a joke and a dictation software mistake. Oops.

[1] most recently https://news.ycombinator.com/item?id=44482876

replies(5): >>44617586 #>>44617610 #>>44617712 #>>44617768 #>>44617817 #
2. Buttons840 ◴[19 Jul 25 17:42 UTC] No.44617586[source]▶
Use two dashes instead of an actual em dash. ChatGPT, at least, cannot do the same--it just can't.
replies(5): >>44617630 #>>44617635 #>>44617694 #>>44617750 #>>44620853 #
3. furyofantares ◴[19 Jul 25 17:45 UTC] No.44617610[source]▶
Maybe I'm misunderstanding - but I don't think LLM's say cow-orkers. Or is that what you mean?
replies(2): >>44617650 #>>44617728 #
4. chaps ◴[19 Jul 25 17:47 UTC] No.44617630[source]▶
As a frequent user of two dashes.. I hate how people now associate it with AI.

Also, that "cow-orkers" doesn't look like AI-generated slop at all..? Just scrolling down a bit shows that most of them are three years and older.

replies(1): >>44617659 #
5. JoshTriplett ◴[19 Jul 25 17:47 UTC] No.44617635[source]▶
Conventionally, in various tools that take plain text as input, two dashes is an en-dash, and three dashes is an em-dash.
6. ◴[19 Jul 25 17:49 UTC] No.44617650[source]▶
7. ◴[19 Jul 25 17:49 UTC] No.44617659{3}[source]▶
8. Jtsummers ◴[19 Jul 25 17:52 UTC] No.44617694[source]▶
It can't use two dashes? Is that like how Data couldn't use contractions (except he did)?
replies(1): >>44621880 #
9. lupusreal ◴[19 Jul 25 17:54 UTC] No.44617712[source]▶
Give away for what, old farts? That link contains a comment citing the jargon file which in turn says that the term is an old Usenet meme.
replies(1): >>44617739 #
10. ffsm8 ◴[19 Jul 25 17:56 UTC] No.44617728[source]▶
As this error seems to be going back a lot longer then LLMs existed (17 yrs), it could be an auto in-correct situation.

Might be incorrectly saved in some spell check software and occasionally rearing it's head

replies(2): >>44617754 #>>44617775 #
11. chaps ◴[19 Jul 25 17:57 UTC] No.44617739[source]▶
Soon HN is going to be flooded with blogs about people trying and failing miserably to find AI signal from noisy online discussions with examples like this one.
12. skeledrew ◴[19 Jul 25 17:58 UTC] No.44617750[source]▶
ChatGPT: "Hold my beer..."
13. chaps ◴[19 Jul 25 17:58 UTC] No.44617754{3}[source]▶
https://ask.metafilter.com/15649/coworkers-why/amp

This goes back a loooooong while.

14. scarface_74 ◴[19 Jul 25 17:59 UTC] No.44617768[source]▶
How is that a “giveaway”? The search turns up results from 7 years ago before LLMs were a thing? More than likely it’s auto correct going astray. I can’t imagine an LLM making that mistake
15. furyofantares ◴[19 Jul 25 18:00 UTC] No.44617775{3}[source]▶
Oh I see the confusion then. It's not an error, it's a joke, and a very old one at that. Like saying Micro$oft.
16. Velorivox ◴[19 Jul 25 18:05 UTC] No.44617817[source]▶
I like to use em-dashes as well (option-shift-hyphen on my macbook). I've seen people try to prompt LLMs to not have em-dashes, I've been in forums where as soon as you type in an em-dash it will block the submit button and tell you not to use AI.

Here's my take: these forums will drive good writers away or at least discourage them, leaving discourses the worse for it. What they really end up saying — "we don't care whether you use an LLM, just remove the damn em-dash" — indicates it's not a forum hosting riveting discussions in the first place.

17. xeonmc ◴[20 Jul 25 00:40 UTC] No.44620853[source]▶
I thought the convention for en-dash is two hyphens straddled between spaces, and three hyphens without spacer for em-dash?
18. Buttons840 ◴[20 Jul 25 04:08 UTC] No.44621880{3}[source]▶
I've asked it before, "please rewrite that but replace the em dashes with double hyphens", and then it says "sure, here you go", and continues to use em dashes.
replies(1): >>44622719 #
19. Ancapistani ◴[20 Jul 25 07:20 UTC] No.44622719{4}[source]▶
Were you using the web interface? If so, that’s likely why. It renders output dynamically on the frontend.

I bet if you did the same through the API, you’d get the results you want.

replies(1): >>44628716 #
20. Buttons840 ◴[20 Jul 25 19:52 UTC] No.44628716{5}[source]▶
Yes, I was using the web interface.