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’re not righn't”
I just say "true" or "my bad."
It’s not just “true”—it’s “my bad.”
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.
I don't even know how to type an em-dash outside latex or where its function differs from a hyphen.
i don't know how to type it in every interface, but as far as function i like the book Eats, Shoots, and Leaves. It's about punctuation.
Your comment is well written but not grammatically correct, since your second sentence is a fragment, and the pronoun “which” has no antecedent.
I read novels all day, and this idea that I’m going to mistake good writing for AI slop is insane. AI follows ugly obvious patterns by mathematical design, which is why despite its enormous progress in other fields (including math), I’ve seen infinitesimal improvement on the creative writing front over the last two years.
Granted, it can trick the average person… who has a 6th grade reading level and scores 20 on the ACT. Yikes.
I politely disagree. Given the metric fuckton of books it ingested in training, I find it hard to believe that it cannot produce something "in the style of" that would pass most cursory examinations.
As for my grammar, I believe that sentence fragments are an acceptable stylistic device when creating emphasis, at least in more informal settings. I would say comment sections qualify, wouldn't you agree?
What you find hard to believe and reality are clearly divergent. And I never said your flawed grammar was bad. My point was to the contrary.
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.
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 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.
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.
Alexandrite.app seems to put it in for you with three hypens --- neat!
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.
I actually use the emdashes — they're convenient on my phone keyboard.
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 above an 8th grade level, you're also in a minority.
You hit the nail on the head! LLM-generated output does not just have some predictable patterns, it's fundamentally different to how two humans would communicate.
Here are some rule of 3 facts about human communication that I just made up by loosely combining first page search results without checking their validity or relevance whatsoever, neatly organized into a list:
Now is the part where I get confused by my own output, so I am now rewriting what I said slightly differently and pretend the semantic meaning changed.
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.
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"?
Pregerant?
Bruh. I immediately noticed so many scams back in the day simply because those were poorly composed with grammar errors.
Kinda fits into the whole image of who’d be doing scams anyway tho - some barely literate shits looking for “easy money”.
That's by design. They weed out the people that are perceptive enough to not fall for the scam instantly, so everyone that follows up is guaranteed to be as dumb as a brick.
Makes sense and seems to be true for many cases.
I’d add tho that I unfortunately met some folks in my life that I’d like to unmeet, and from what I’ve perceived - people doing scams are also far from being sharpest tools in the shed.
Not to mention that back when scam calls on the phone were still quite common - wasting time of those guys was one of the funniest things to do. Like, at some point that person should’ve realised I’m shitting them, just have to make them move off the script and watch them come up with bullshit that truly reflects that person’s stupidity.
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
So, i live near the worst college. Absolute worst on every ranking (if it even gets ranked) except the parking lot is nice. Let's pretend it's called [Lemmy Junior College].
so i was getting to know some people once. mostly because i liked this gal and they were her friends. we were talking and i used a three syllable word. one of them stopped me and said "slow down, this ain't [Lemmy Junior College]". worst part is, they weren't joking.
my wife and i now say that to each other, generally to say i understand but in a joking manner.
three fucking syllables. seriously. i wish i remembered which one. it was probably syllable. it wasn't like i was trying to gossip about those dang heteroskedastics.
Also us crossword-heads
you word good, dude. i'ma start using literature as a singular too
Whoa, holy shit, what a mindfuck
There's... more of me here?
jokes on you, I learned about texting forensics on a true crime documentary when I was 10 and have purposefully made my writing style unique since then.
also noone is written as one word, fight me.
rip fingerprinting and sentiment analysis
but fr that's really cool, do you remember the doc name? i want to learn as well
The problem with writing "no-one" like that is that English doesn't have a way of differentiating between the "oo" sound and "o-o" besides hyphenating or adding a space. I've always hated the fact that "cooperation" is spelled that way for a similar reason.
Actually, we did, it's just mostly been abandoned because "Accent marks scary," or something. If you look in older texts, you can find words written like coördination, coöperation, etc, to distinguish between two identical short vowels, meant to be pronounced separately from the long vowel in words like soon. I want to say the New Yorker, or some similar periodical still prescribes its use in articles they publish, but that's about it these days.
Interesting. How recently were those accents used? I haven't really studied much English outside of modern use, but I seem to remember middle and old English having them, only for them to vanish in "modern" use.
The Wiktionary page cites Webster's Revised Unabridged Dictionary from 1913, so it survived at least until the early 1900s. I seem to recall coming across a similar use of diaeresis in novels from the 1800s.
Because of the inevitable "dont worry mom, im fine dont look for me" text your killer will send?
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.
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.
@lemmy.world
No trolling, bigotry or other insulting / annoying behaviour
This is non-politics community. For political memes please go to !politicalmemes@lemmy.world
Check for reposts when posting a meme, you can only repost after 1 month
No bots without the express approval of the mods or the admins
No advertisements or spam. This is an instance rule and the only way to live. We also consider AI slop to be spam in this community and is subject to removal.
A collection of some classic Lemmy memes for your enjoyment
go to feed...
@lemmy.world
No trolling, bigotry or other insulting / annoying behaviour
This is non-politics community. For political memes please go to !politicalmemes@lemmy.world
Check for reposts when posting a meme, you can only repost after 1 month
No bots without the express approval of the mods or the admins
No advertisements or spam. This is an instance rule and the only way to live. We also consider AI slop to be spam in this community and is subject to removal.
A collection of some classic Lemmy memes for your enjoyment
go to feed...
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
save