Live data from Hacker News

New accounts on HN more likely to use em-dashes

marginalia.nu

561–570 of 643 posts

Re: New accounts on HN more likely to use em-dashes

#561
post #554

Earlier quoted context omitted.

You can make that statement without subjecting people to slop. Edit: I realize that sounds harsh. Not trying to be. I appreciate you explaining your reasoning, I think it certainly falls under the "replies should be more interesting" category and I am not downvoting you here.

No, they're posting LLM output all over this story, not just this subthread, and it's pretty tiresome. edit: he only did it twice, I exaggerated and that's my bad.

> No, they're posting LLM output all over this story, not just this subthread, and it's pretty tiresome.

Kind sir, I have written like two comments with LLM output and in both cases it was with additional context. [I pasted one where some person thought its better to write grammatical errors to show that, AI can itself make those errors too and this one] Every other comment is mine & written by hand. (or well one comment was written by voice with handy that people recommended here :D)

Now there's a point you can make if my writing can be sloppy and I totally would get that but sometimes I get over-enthusiastic about a particular topic.

This comment, I made weeks ago seem apt for me to use here and please don't mind if I use the same right now as well: https://news.ycombinator.com/reply?id=46986446

I think I only tried to reference LLM in ironical situations in both the times that I shared or atleast so were my intentions. Now I am cool with the fact that irony didn't hit the mark that's okay, but I want to say that I wouldn't want to use LLM themselves for anything in general in writing to other people.

Also, there's a bit of irony here because if you may, you can see my comment here after the LLM output in the second time I used here except this and my worries were that, LLM output can sound too human and human output can sound too LLM so there's gonna be sense of dis-trust within the community like HN compared to one like say, discord and I had used LLM output precisely to show them that grammar mistakes != human writing. [https://news.ycombinator.com/reply?id=47157571]

Sir, to give you context, Do you really think that I am gonna be using LLM to unironically write my messages?, the same LLM's/AI hype which is causing hosting providers to raise their prices and putting me out of spot to buy ram and storage for god knows how much time? If that's the case, I hope you can know what my priorities are.

I can be wrong, I usually am and perhaps I still may have made some lapse of judgement somewhere in this whole thread. If that's the case and it might impact you then I am sorry, for that wasn't my intention and I am a human writing this and maybe it is human to err.

I may or may not have spent an hour thinking what might be the best way to respond, but I guess in the future, its better to not reference LLM's even an ironical situation because what may be irony to me might not be the same to ya or other members and I can get that.

Do you know what the real irony is right now? Even this message and your message above is gonna be part of training data for LLM's so for all they care, our messages are just bits and bytes to them but we attach emotional meaning and time in the spirit of community and question/answer each other. LLM's are so baked in irony that its the tower of bable of irony.

Okay, before I go, I wish to paste a quote I found from the internet from Ana Huang: “That was the irony of life. People always reminisced about the good old days, but we never appreciated living in those days until they were gone.”

[Source: https://quotefancy.com/quote/4027241/Ana-Huang-That-was-the-...]

Re: New accounts on HN more likely to use em-dashes

#562
post #554

Earlier quoted context omitted.

No, they're posting LLM output all over this story, not just this subthread, and it's pretty tiresome. edit: he only did it twice, I exaggerated and that's my bad.

> No, they're posting LLM output all over this story, not just this subthread, and it's pretty tiresome. Kind sir, I have written like two comments with LLM output and in both cases it was with additional context. [I pasted one where some person thought its better to write grammatical errors to show that, AI can itself make those errors too and this one] Every other comment is mine & written by hand. (or well one com…

You're right, you posted a lot about LLM style but only pasted LLM output twice. I apologize for misrepresenting your posting in that fashion.

I do think you would do well to revisit the thread you linked at https://news.ycombinator.com/reply?id=46986446, because I saw the OP's comment when it was posted, I agreed with it then and I kind of still do.

Re: New accounts on HN more likely to use em-dashes

#564

Earlier quoted context omitted.

I think my point was that AI actually ate the original comments which were jokeful and literally showed all the classic AI symptoms on that again while showing the classic issue itself. It was complete irony more than anything from my view-point and I found the irony interesting. The "interesting" thing about this is that you can give any rudiculous idea to AI and say autocomplete after this: "You are absolutely righ…

You can make that statement without subjecting people to slop. Edit: I realize that sounds harsh. Not trying to be. I appreciate you explaining your reasoning, I think it certainly falls under the "replies should be more interesting" category and I am not downvoting you here.

Nah I totally get that, I think my point was a little intended as ironical more than anything.

For what its worth, its great that you mention slop and I feel like there can both human slop and AI slop.

Had to look up cambridge for definition of slop there but slop in this context means, content on the internet that is of very low quality, especially when it is created by artificial intelligence:

Quality essentially sums down to being "good" whose definition is "very satisfactory, enjoyable, pleasant, or interesting"

I guess in retrospect, My comment can be considered unsatisfactory/less-interesting as you mention as well, that can be totally true.

I guess I can (try?) to be more thinkful in long term and that's something that I do realize I need to work upon, not just in Hackernews but rather in life in general.

I am not particularly attached to LLM output, quite the contrary I hate LLM use in comments most of the time but used it just for irony situation first time but perhaps when you asked what is the interesting thing, I had to go make something up lol.

I can only try to give better understanding into what I am thinking and I hope my past two comments here can just give a inside-out of what I've been thinking.

Have a nice day.

[Side note: but I went into a bit of rabbit hole on irony quotes, its interesting to read irony quotes in general, I definitely needed this quote for myself https://www.azquotes.com/quote/379798?ref=irony, not sure why its in irony section tho. But yea]

Re: New accounts on HN more likely to use em-dashes

#565

Prior to the rise of LLM-written posts and the natural reaction of hair-trigger suspicion, I used to em and en dash fairly often in posts on HN. No reason really other than being a bit of a typography geek who happens to have always used dashes in casual writing instead of semicolons. So when I was setting up a modifier-key keyboard layer with AHK many years ago I put the em dash on modifier+dash just because I could…

My teenager recently asked me why I write like a chatbot, apparently unaware that some human beings prefer to write in complete sentences with attention to details like spelling, punctuation, grammar, and capitalization, and that LLMs were trained on this sort of writing. This makes me think of the fad where people on youtube will hold a microphone up in frame, because it somehow connotes authenticity. I'm sure some…

I've been thinking about the "human signal" ever since the first GPT. Interestingly, a few years ago I would get people questioning whether my posts were written by GPT, but that's stopped now. I'm not sure if it's because they no longer think that or because it's not interesting any more.

I wonder if there's a way we can communicate that LLMs fundamentally can't keep up with. If LLMs have hit on being exactly the way our brains work then I guess not. But maybe we still have something special. I haven't tried how well LLMs understand language written like in Iain M. Banks's Feersum Endjinn.

Re: New accounts on HN more likely to use em-dashes

#566

This feels like an existential threat to HN, and to the general concept of anonymous online discourse. Trust in the platform is foundational, and without it the whole thing falls down. Requiring proof of identity is the only solution I can think of, despite how unappealing it is. And even then, you'll still have people handing their account over to an LLM. I really struggle to imagine a way around it. It could be tha…

Another option instead of using identity is to use proof of work or hashcash such that anyone who thinks a comment is valuable can use some hash rate to upvote it. It doesn't matter how the content was generated, only that someone thought it was important, and you can independently verify this by checking how much hash effort went into hashing for that comment. This also does not require any identity either.

Advertisers are more willing to spend money to promote content than an individual is willing to do the same...

Re: New accounts on HN more likely to use em-dashes

#568
post #562

Earlier quoted context omitted.

> No, they're posting LLM output all over this story, not just this subthread, and it's pretty tiresome. Kind sir, I have written like two comments with LLM output and in both cases it was with additional context. [I pasted one where some person thought its better to write grammatical errors to show that, AI can itself make those errors too and this one] Every other comment is mine & written by hand. (or well one com…

You're right, you posted a lot about LLM style but only pasted LLM output twice. I apologize for misrepresenting your posting in that fashion. I do think you would do well to revisit the thread you linked at https://news.ycombinator.com/reply?id=46986446 , because I saw the OP's comment when it was posted, I agreed with it then and I kind of still do.

> You're right, you posted a lot about LLM style but only pasted LLM output twice. I apologize for misrepresenting your posting in that fashion.

Thanks for the apology, I appreciate it.

> I do think you would do well to revisit the thread you linked at https://news.ycombinator.com/reply?id=46986446, because I saw the OP's comment when it was posted, I agreed with it then and I kind of still do.

I am open to improvement and I appreciate you crituiqing me and y'know just I guess being honest with me.

I am gonna be honest with ya as well, I can't guarantee this overnight.

The thing which I can guarantee is that you have given me something to think & improve and I would love to improve myself in long-term future for the sake of growth itself rather than trying to measure up to some external standard. Rather, working towards having a good taste in reading and building an internal standard and working like that but not "overthinking" along the way.

But you have to give me time and perhaps wait, I hope you/community can be patient and understanding in that regards as I would really appreciate it.

Thanks, Have a nice day.

Re: New accounts on HN more likely to use em-dashes

#569
post #494

Earlier quoted context omitted.

My teenager recently asked me why I write like a chatbot, apparently unaware that some human beings prefer to write in complete sentences with attention to details like spelling, punctuation, grammar, and capitalization, and that LLMs were trained on this sort of writing. This makes me think of the fad where people on youtube will hold a microphone up in frame, because it somehow connotes authenticity. I'm sure some…

Idiocracy called this "talking like a fag".

“Welcome to Costco. I love you.”

Re: New accounts on HN more likely to use em-dashes

#570

Fwiw I did some more comparisons, looking for words disproportionately favored by noob comments: word noob new p-value ---------------------------- ai 14.93% 7.87% p=0.00016 actually 12.53% 5.34% p=1.1e-05 code 11.47% 6.04% p=0.00081 real 10.93% 2.95% p=2.6e-08 built 10.93% 2.11% p=2.1e-10 data 8.93% 3.51% p=6.1e-05 tools 7.6% 2.67% p=5.5e-05 agent 7.47% 2.95% p=0.00024 app 7.2% 3.09% p=0.00078 tool 6.8% 1.83% p=8.5e…

Worth pointing out that calculating p-values on a wide set of metrics and selecting for those under $threshold (called p-hacking) is not statistically sound - who cares, we are not an academic journal, but a pill of knowledge. The idea is, since data has a ~1/20 chance of having a p @OP have you considered calculating Cohen's effect size? p only tells us that, given the magnitude of the differences and the number of…

> The idea is, since data has a ~1/20 chance of having a p Are you saying p is uniformly distributed over any data set? That doesn't jive with my limited understanding of entropy. What's this based on?
Post reply on HN