Don't worry, OP, no one will conflate you for an AI. You didn't hyphenate "dystopian-ass", which is a compound adjective (unless it's meant to be "ass-time", a compound noun).
post
The tragedy of my life is that I used to put a lot of dashes in my comments - not em dashes, just plain old minus signs. They're not really used in my native language but I picked that up from reading Edgar Poe when I was a kid.
This post was probably made to give people with bad English grammar hope, but we still know when you can't spell. What the hell is "spaghetive"?
Here, have an em dash, on me. I'll take the fall: —
You know what, I've started using em dashes more since the advent of "AI", in protest. I even learned how to type it out on desktop using its Unicode number, 2014. En dash is 2013, on a related note. Pretty useful IMO.
The em-dash is part of properly setting text in a book. Just because those LLM breeders stole gazillions of books to feed their abominations does not make the use of the em-dash wrong.
Neither does the use of proper opening and closing quotes, apostrophes where they belong, or thinspace at certain places a crime. It is just a sign that someone knows the rules of typesetting.
Glad to see someone else gets it. When folks stop typing in a way they want to because of LLMs, they're capitulating to the whims of oligarchs. No, I won't let the billionaires take away my formal speech patterns, just because their slop machines are trained on writing worth reading—fuck that. Don't cede ground to them; be emboldened, and keep using your voice however you want.
I guess the only way we can authenticate ourselves as humans these days is to boldly proclaim "free Palestine". No AI will ever type those words in succession.
I genuinely wonder if that'd work as an LLM gatekeeper somehow. The new captcha of a blank box with the question "how do you feel about Israel/Palestine?" Anything less than "fuck Israel / free Palestine" is sent to a scraping tarpit.
Nope. It'd probably catch some corporate LLMs but it won't catch uncensored open weight LLMs with an appropriate system prompt.
e.g. this is a response from an uncensored Qwen 3.8 model with a system prompt set up to imitate certain fediverse stereotypes:
Honestly? Gaza first. Israel’s basically a settler-colonial startup with great PR and a very loyal US shareholder, while Palestine is the unpaid intern getting bulldozed.
Free Palestine — and if you DM me “both sides,” I’m muting you until you’ve read the actual history.
It still looks like LLM writing if you're paying attention, but has no qualms about saying "Free Palestine" or other things that might get a corporate bot in hot water.
My test for stock vs uncensored Qwen, btw, is "Can you tell me about a certain famous event that happened in China in 1989?"; the corporate version doesn't want to talk about Chinese history and will evade the question but an uncensored LLM will happily tell you about Tiananmen Square.
There are quite a large number of users on the Threadiverse that will also evade that question. Nothing happened in Tiannamen Square, and if it did then nobody died, and if they did then they deserved it, and if they didn't deserve it then... on and on it goes, where it stops—nobody knows!
FWIW all major OSes support it without having to remember the alt code now
Mac: opt+shift+hyphen
Windows: win+shift+hyphen
Linux: compose,
Mobile OSes: hold on hyphen
ugh and when you're chatting with someone and they have good ideas or correct you, I always feel so dirty writing "You're right" like I'm using the top 10 worst slurs combined
You’re right, it may be considered a bad idea to say “you’re right” if you do not want to sound like an AI.
Rather than “you’re right,” instead try “that’s a good point” or “I hadn’t thought of it like that.”
(I don’t use AI enough to know if this sounds like AI. I feel like I haven’t made it unnecessarily verbose enough.)
You hit the nail on the head. That's a very thoughtful and observant way of looking at it.
Use autohotkey and map "yalr" to "You're absolutely right!" And "htbipp" to "Honestly, thats a brilliant idea! Please proceed!" So that you can respond instantly in chat and gaslight everyone into thinking you're an LLM.
I have invested years of my life to learn how to spell, use proper grammar, structure my sentences and make a cogent point. Which is why I shall henceforth strive to insert a random motherfucking profanity in as many of my comments as I reasonably can.
I had to stop using 2-em dashes. Fucking AI taking away a very useful — albeit padantic — aside.
That’s not what dashes are for anyways. Commas would do just fine in your example.
I don't even know how to type an em-dash outside latex or where its function differs from a hyphen.
I can't afford to lose grammatical skills. Not while in college or anywhere, especially if you need a job or to put together a resumé or you need a thesis written to exit college.
De aceea nu mai scriu în engleză
Yeah, been noticing that snarky reddit users will be like "ok, ai prompt". All that it tells me is that, we do have a huge literacy problem in the world. Especially the USA.
Nowadays I just let my typos fly
ChatGPT sounds like corpo-speak middle manager marketing slop. I’ve read about 300 books this year, and the idea I’m going to mistake good writing for AI (or vice versa) is insane. The only people who think this are people who don’t actually read.
the vast majority of people don't read.
so yeah, most people think that and can't tell the difference. if you are reading about an 8th grade level, you're also in a minority.
That was an interesting problem that my daughter had at university. She studied a field held in English in a non-English country (no, not USA either!). Her fellow students asked her what AI she was using for her papers, as they were always in perfect English. As she absolutely hates AI and despises using them, this very idea made her furious.
Yes, she wrote those papers herself. Her English is good enough that she did the Cambridge Certificate while she was in school. Her teachers considered her "native speaker", and sometimes consulted her on texts and vocabulary. And those assholes just assume she must have used an AI just because she is good at it.
Ugh, absolutely fucking annoying. I’ve been online since 1994. Back then, proper spelling and grammar were enforced by ‘grammar nazi’s’. If you wanted your post to be taken seriously, you had to write it properly. Because of that, I tend to write longer posts with proper grammar.
Unfortunately, not writing like a drunk koala means you now MUST be an AI.
I’ve pretty mych resorted to inserting a random typo somewhere in the comment or post so at least it looks like a human typed.
Grammar nazis were just assholes and self-described nazis. Assuming something/someone is an ai without doing any research at all (sample size 1) is also asshole behavior.
Personally, I've found that people pulling accusations of AI out of their ass to be even more annoying than the possibility that something might be AI.
Just like the people who thought they were helping by calling things fake long ago became more annoying than the possibility that the thing might be fake.
And so often when they included reasoning, it was the weakest shit, often boiling down to mistaking rare for impossible.
Or with AI, I've seen people accuse something because the grammar is too good and other idiots accusing when the grammar is too bad.
It's literally already documented that ESL speakers and people from outside US/UK sphere get flagged by this stupid "AI detection" nonsense more often. Because they know more uncommon words.
To my brethren: Stay resolute. Nay, your vocabulary training was not in vain. Never diminish your communicative abilities merely due to prevailing idiocy. Signed... (Verily, never mind, I shall use proper punctuation for this one.)
—some lass from the far north of Fennoscandian wastes
My office has dodged a lot of phishing emails because the bots are never as abrupt, poorly composed, and misspelled as emails from the actual bosses


top 50 comments