Hacker News Reader

Comment by chinathrow | Hacker News Reader

The '—' gave it away. No one types this character on purpose.

I really loved how easy MacOS made these (option+hypen for en, with shift for em), so I used to use them all the time. I'm a bit miffed by good typography now being an AI smell.

On MacOS (and I have this disabled since I'm not infrequently typing code and getting an — where I specced a - can be not fun to debug)...

Right click in the text box, and select "Substitutions". Smart dashes will replace -- with — when typed that way. It can also do smart quotes to make them curly... which is even worse for code.

(turning those on...)

It is disappointing that proper typography is a sign of AI influence… (wait, that’s option semicolon? Things you learn) though I think part of it is that humans haven’t cared about proper typography in the past.

Just because you don’t, doesn’t mean other people don’t. Plenty of real humans use emdash. You probably don’t realise that on some platforms it’s easy to type an emdash.

In Office apps on Windows just type two hyphens and then a word afterwards and it will autoconvert to an em-dash.

iOS keyboard as well.

And where did you suppose AIs learned this, if not from us?

Turns out lots of us use dashes — and semicolons! And the word “the”! — and we’re going to stuff just because others don’t like punctuation.

Books use it more liberally, internet writings not so much. Also some languages are much more prone to using it while some practically never use it

I'm starting to wonder if there's a real difference between the populations who use em dashes and those who think it's a sign of AI. The former are the ones who write useful stuff online, which the AIs were trained on, and the latter are the consumers who probably never paid attention to typography and only started commenting on dashes after they became a meme on LinkedIn.

Holy ego lol.

Indeed.

I find it disturbing that many people don't seem to realize that chatbot output is forced into a strict format that it fills in recursively, because the patterns that LLMs recognize are no longer than a few paragraphs. Chatbots are choosing response templates based on the type of response that is being given. Many of those templates include unordered lists, and the unordered list marker that they chose was the em-dash.

If a chatbot had to write freely, it would be word salad by the end of the length of the average chatbot response. Even its "free" templates are templates (I'm sure stolen from the standard essay writing guides), and the last paragraph is always a call to further engagement.

Chatbots are tightly designed dopamine dispensers.

edit: even weirder is people who think they use em-dashes at the rate of chatbots (they don't) even thinking that what they read on the web uses em-dashes at the rate of chatbots (it doesn't.) Oh, maybe in print? No, chatbots use them more than even Spanish writing, and they use em-dashes for quotation marks. It's just the format. I'm sure they regret it, but what are they going to replace them with? Asterisks or en-dashes? Maybe emoticons.

Do you have a pointer to documentation on that, or a keyword to google? Would like to find out more.

All that may be true. Let’s assume for argument that it is. I’ve had people call out my own handwritten, zero-AI comments (which are 100% of them) as likely to be AI because I used proper grammar, common punctuation, and a bullet list.

To me, “ah ha, gotcha, AI wrote this!” comments are more common and tedious than the AI-augmented comments themselves.

The AI is trained on human input. It uses the dash because humans did.

Yeah but a dash, at least on my keyboard is a '-', not the one quoted above.

If I type two dashes—like this—my phone changes it into a special character. Same for three dots…

En and em dashes are easily accessible on both my laptop's and phone's keyboard layouts and I like using them, just like putting the ö in coöperate. It's sad if this now makes me look like a robot and I have to use the wrong dashes to be more "human".

TIL that some people spell cooperate with an "ö".

As a Swedish native it really breaks my reading of an English word, but apparently it's supposed to indicate that you should pronounce each "o" separately. Language is fun.

As a native English speaker, it also breaks my reading of "cooperate". Never seen it before. I think parent is just annoyingly eccentric for the sake of it.

Most commonly seen in naïve, and the New Yorker

I admit that latter part is just for whimsy, because I think it looks fun. The dashes I like for their aesthetics and if that makes me eccentric then so be it. They shouldn't distract anyone's reading, or at least they didn't use to before LLMs.

Using umlauts to signal that a vowel is pronounced separately is common in a number of languages (like Dutch).

Yeah, I know.

It's just confusing for us poor Swedes since "ö" in Swedish is a separate letter with its own pronunciation, and not a somehow-modified "o". Always takes an extra couple of seconds to remember how "Motörhead" is supposed to be said. :)

That kind of use technically makes it a diaeresis, not an umlaut.

But it's not used as an Umlaut here, that's exactly what's confusing. Here this is used as a trema/diaeresis.

Em dashes are widely used. The diaeresis is only used in The New Yorker and those that copied their style.

If you’re using the dash on your keyboard (which is a “hyphen–minus” character) in place of a en dash or em dash, then you are using the wrong character. That’s fine — it’s certainly more convenient, and I wouldn’t call you out on it — but it’s silly to assume that other people don’t use the correct characters.

https://www.grammarly.com/blog/punctuation-capitalization/da...

I'm skeptical this is the reason:

- Chatgpt uses mdashes in basically every answer, while on average humans don't (the average user might not even be aware it exists)

- if the preference for em dashes came from the training set, other AIs would show the same bias (gemini and Le chat don't seem to use them at all)

> Chatgpt uses mdashes in basically every answer, while on average humans don't

I would not be shocked if an aspect to training is bucketing "this is an example of good writing style" into a specific category. Published books - far more likely to have had an editor sprinkle in fancy stuff - may be weightier for some aspects.

My iPhone converts -- to — automatically. So does Google Docs / Gmail (althought I'm not certain if that's on their end or my Mac's auto-correct kicking in). Plenty of them out there.

> other AIs would show the same bias

Unless they've been trained not to use it, now that a bunch of non-technical people believe "emdash = AI, always".

Is that why it uses colorful emoticons, too? Was it trained on Onlyfans updates?

It was trained on everything they could get their hands on.

Yes, it uses emoticons because human writers sometimes use emoticons.

Or at least not anymore since this became the number 1 sign whether a text was written with AI. Which is a bit sad imo.

I do all the time, but might have to stop. Same with `…`.

Don't let them win. Stand proud with your "–" and your "—" and your "…" and your "×".

I dislike the ellipsis character on its own merits, honestly. Too scrunched-up, I think - ellipses in print are usually much wider, which looks better to me, and three periods approximates that more closely than the Unicode ellipsis.

In the words of Michael Bolton, "Why should I change? He's the one who sucks."

That got a giggle out of me. Not entirely relevant but AI tends to be overzealous in its use of emojis and punctuation, in a way people almost never do (too cumbersome on desktop where majority of typing work is done)

https://www.scottsmitelli.com/articles/em-dash-tool/

I started using hyphens a few years ago. But now I had to stop, because AI ruined it :(

Academia certainly does, although, humorously, we also have professors making the same proclamation you do, while while en or em dashes in their syllabi.

Keep in mind that now that people know what to pay attention to: em-dash, emojis, etc. they will instruct the LLM to not use that, so yeah.

Two dashes on the Mac or iOS do it unless you explicitly disable it, I think.

I absolutely bloody do -- though more commonly as a double dash when not at the keyboard -- and I'm so mad it was cargo-culted into the slop machines as a superficial signifier of literacy.

I used to.

Reply on news.ycombinator.com