Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

To quote Office Space, “Why should I change? He’s the one who sucks.”


Mostly because when I see an em dash now, I assume that it was written by AI, not that the author is one of the people who puts enough effort into their product that they intentionally use specific sized dashes.

AI might suck, but if the author doesn't change, they get categorized as a lazy AI user, unless the rest of their writing is so spectacular that it's obvious an AI didn't write it.

My personal situation is fine though. AI writing usually has better sentence structure, so it's pretty easy (to me at least) to distinguish my own writing from AI because I have run-on sentences and too many commas. Nobody will ever confuse me with a lazy AI user, I'm just plain bad at writing.


> assume

There's your trouble. The real problem is that most internet users are setting their baseline for "standard issue human writing" at exactly the level they themselves write. The problem is that more and more people do not draw a line between casual/professional writing, and as such balk at very normal professional writing as potentially AI-driven.

Blame OS developers for making it easy—SO easy!—to add all manner of special characters while typing if you wish, but the use of those characters, once they were within easy reach, grew well before AI writing became a widespread thing. If it hadn't, would AI be using it so much now?


As someone who frequently posts online- with em dashes- I wonder if I am part of the problem with training llms to use them so much- and am going to get punished in the future for doing so.

I also tend to way overuse parenthesis (because I tend to wander in the middle of sentences) but they haven't shown up much in llms so /shrug.


If you’re judging my writing so shallowly, I don’t think I’m writing for you.


> If you’re judging my writing so shallowly, I don’t think I’m writing for you.

No, you are writing for people who see LLM-signals and read on anyway.

Not sure that that's a win for you.


"Seeing LLM-signals" == "reading shallowly", so I think I covered that case.


Or you're writing for the people who haven't deluded themselves into thinking that they're magical LLM detectors, which definitely does seem like a win.


> Or you're writing for the people who haven't deluded themselves into thinking that they're magical LLM detectors, which definitely does seem like a win.

What delusion? The false positive rate just on HN alone is so low it's not even a rounding error.


I don’t think I’m judging shallowly- there is no em-dash on a standard keyboard. The one way it ends up in real writing is if you use a typesetting program like LaTeX, or Word changes an en-dash with auto formatting, or the user consciously interrupts their writing flow to insert the character with a special keystroke combination or by pasting it in. The proportion of people who do any of those things in writing for the web is quite small. The number of clearly AI written posts with em-dashes is quite large. So large, that I immediately suspect AI writing when I see an em-dash and I rarely see countering evidence that suggests the author is human but meticulous about how they write.


> there is no em-dash on a standard keyboard. The one way it ends up in real writing is (…)

Then you proceed to list multiple ways to do it, but neglected to mention that by default on Apple operating systems they are inserted automatically when typing “--“. It’s something you have to explicitly turn off of you don’t want it. On Apple mobile operating systems you can also long press the hyphen to get the option. Em-dashes are trivial to type.


Both of the examples you gave both fall under "a special keystroke combination," which I did list. Typing "--" is two keystrokes compared to one for an en-dash.

The iOS example isn't just "long press the hyphen" it's "press the [123] button, long press the hyphen, and slide your finger over the em-dash" compared to "press the [123] button, long press the hyphen" for the en-dash.

If you're going to argue at least be genuine. I didn't say it was hard to type an em-dash, I showed that every way to get an em-dash into your writing takes an extra step. Taking an extra step compared to other characters means it isn't trivial.

For someone writing publication quality work, em-dashes appear and if I see an em-dash in a book I don't assume AI writing. But for comments on the internet or a blog posts that aren't meticulous everywhere else, an en-dash is a pretty good signal that the work is AI generated. When people are writing, needing an extra step to insert an em-dash is disruptive to most people's train of thought.


> I didn't say it was hard to type an em-dash

Neither did I say you said that. I only said they are trivial to type. Which they are. I do it all the time, and it doesn’t interrupt my train of thought any more than a comma. I also do the keyboard shortcuts for things like “smart quotes” and apostrophes (’). For some of those I even have my own special snippets in Alfred, like typing "" produces “” with the caret in between. I can’t even tell you what the exact shortcuts for those are without looking at my fingers, because they are so ingrained in my muscle memory. I know I’m far from alone in that.

> But for comments on the internet or a blog posts that aren't meticulous everywhere else, an en-dash is a pretty good signal that the work is AI generated.

Provably false.

https://news.ycombinator.com/item?id=45071722


To be fair, if they don’t know that they probably run Android, and are you even writing for them?

I bet their bubbles are… green. The horror!


To continue the story, the guy saying this got fired and probably wouldn't have without taking this stand.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: