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 people are already embracing a bit of sloppiness in their writing as a signal of humanity; I'm equally sure that future chatbots will learn to do the same.

> default to pretty good punctuation, spelling and grammar

If leaving out the Oxford comma here was an intentional joke I both commend and curse you!

You're absolutely right! I kid. I'm also a former avid user of the em-dash, but have mostly stopped using it. I've even started replacing em-dash usage with commas, which often results in a slightly awkward, perhaps incorrect, but quaintly artisanal sentence with a LaCroix-like spritz of authenticity.

My double-space-after-a-period though, I will keep that until the end. Even if it often doesn't even render in HTML output, I feel a nostalgic connection to my 1993 high school typing teacher's insistence that a sentence must be allowed to breathe.

Have the same problem but with bullet points, which I learned to type years ago and have used on HN for a long time:

• Like

• This

(option-8 on a Mac US keyboard layout). Now it looks like something only an LLM would do.

Ex-academic here. I too use/tended to use em-dashes quite a bit. It's easy to compose in Linux (Gnome) with a real keyboard: Ctrl Shift U 2014 is ingrained in my head from using them all the time in my academic work.

Were you using them as a replacement for a comma--without spaces on both sides of the em-dash--like how I did just now? If no, you are safe from being mistaken for an LLM program. Honestly, while it is a legitimate punctuation rule, I've never seen a human on the internet to write like that. But LLMs do it constantly, whenever they generate long enough sentences.

How dire the literacy crisis, that chatbots are their only exposure to composition.

I also used em-dash before LLMs, though I would not call myself a typography geek. But yesterday I wrote a birthday message to someone and replaced my em-dashes with minus signs, because I did not want them to think that my message is LLM generated..

I do agree... I sometimes use worse grammar (like that ellipses) and leave in typos just so my comments feel more "real" now.

> Now someone may search old posts without a time cutoff and assume I'm an LLM.

I use em dashes, and I don't care whether or not someone assumes I'm an LLM. Typography exists for a reason.

When every breath is a Turing test, AmIBotOrNot?

I’m waiting for a Philip K. Dick bot to declare me non-human.

Am I the only one who in a Captcha test sometimes wants a different option for the “I am Human” check box? Ironically really since to prove we’re human we have to check the boxes with a crossing in them, no account to be made of people who call them zebra crossings.

Fwiw your comment has lots of human tells and doesn't seem AI generated at all.

My rage–induced habit of ignoring typos caused by the iPhone autocorrect and general abuse of English is suddenly authentic and not lazy and slightly obnoxious (ok maybe it's still those things too)

>I put the em dash on modifier+dash

This is the default on Macs

I'm also increasingly aware that my own writing style and punctuation seem to line up with what might be associated with an AI, but some of the tells (em-dashes, spaces after periods, etc) seem like artifacts of when in history we learned to write.

I wonder how much crossover there would be between a trained text analysis model looking for Gen-X authors and another looking for LLM's.

It really is unfortunate that such a fun piece of punctuation has been effectively gutted. This isn't even really limited to just the em-dash, but I don't know if there's another example of a corporation (or set of them) having such a massive impact on grammar and writing as OpenAI and their ilk have.

Entire sentence structures have been effectively blacklisted from use. It's repulsive.

I have consistently used em-dashes, either in the form of alt+- on MacOS, or in the form of `--` in LaTeX (or `---`), for the last 30 odd years.

Now I find myself deliberately making things worse to avoid being accused of not being human! Bah!

I do a similar thing — also with AHK! — and I don’t intend to stop. I think probably the AI/LLM bubble will pop before I consider changing my habits there.

Tip: Patterns like “It’s not just X, it’s Y” are a more telltale sign of LLM slop. I assume they probably trained on too much marketing blurb at some point and now it’s stuck.

Exactly what an LLM would say, haha.

I use “-“ because I thought the amount of parentheticals I was using was a bit unhinged. In these times of TLDR, I sometimes move the aside to the bottom as an afterthought instead of leaving it inline.

I dunno this en versus em dash stuff, I just use the minus sign on my keyboard.

Nice try

ChatGPT evolves, everything grows. In AI speech, tells abound. Comma, emphasis. A new way, a better way.

I also used — and "proper" quotes which macOS/iOS puts in for you anyway

I also like …

This is like ruining swastikas and loading rainbows

I used to do that too… even using the ellipses character instead of three dots. But on the other hand I'm not a native English speaker and have poor spelling (i.e. words pass spell check, but are incorrect).

That's one of the signals I use to detect if YouTube videos are AI slop. If it's narrated by a non-native speaker, it's much more likely to be high quality. If it's narrated by a British voice with a deep timber, it's 100% AI.

[dead]