No, virtually nobody used the em dash before AI.
I will not be psyopped.
I will not be psyopped.
Timeline
Post
Remote status
Replies
32
@WandererUber you're absolutely right!
@WandererUber it happens a lot in corpo and academic settings where people used microsoft word and it automatically changed dashes to other dashes like em dashes depending on context (also some niggas used to use them on purpose even without word (I was niggas (Ive gone out of my way to break that habit now though)))
@wertimer dude this fucking shit is getting so out of hand. Marked content is probably 100% hallucinated. There is no way that this is true and the entire paragraph is wasted space.
Serves me right for expecting people not to overuse AI in an AI sub, I guess, but still. This is so bad for signal-to-noise and no one is realizing it. I wonder how non-chuds (or sleeper chuds) deal with this garbage. This is unreadable. Their monthly roundup of interesting models has totally incorrect summaries for most of them, and when you click through to the github repo readmes, those are ALSO written by AI and have these idiotic explanations everywhere.
sorry had to vent
Serves me right for expecting people not to overuse AI in an AI sub, I guess, but still. This is so bad for signal-to-noise and no one is realizing it. I wonder how non-chuds (or sleeper chuds) deal with this garbage. This is unreadable. Their monthly roundup of interesting models has totally incorrect summaries for most of them, and when you click through to the github repo readmes, those are ALSO written by AI and have these idiotic explanations everywhere.
sorry had to vent
@karna fair enough, but then they would show up in Word documents etc, not on websites though. No one was writing github readmes or reddit posts in Word.
@WandererUber re reddit, maybe they wrote drafts in word. I've written drafts of longer poasts in vim... (agreed on the readmes though)
@WandererUber academicfags used them
LLM output is a combination of middle manager speak and academic fart huffing
LLM output is a combination of middle manager speak and academic fart huffing
@bronze no they didn't
@bronze maybe you mean arts graduates or something, definitely not in use in STEM
@WandererUber maybe
i'd like to find the guy who popularized "its not about x, its about y" and strangle him to death
i'd like to find the guy who popularized "its not about x, its about y" and strangle him to death
@WandererUber its even worse than the em dash imo
i hear it everywhere now, i know whatever video im watching is AI when its doing a comparison every other paragraph to emphasize whatever point the prompter was trying to make
i hear it everywhere now, i know whatever video im watching is AI when its doing a comparison every other paragraph to emphasize whatever point the prompter was trying to make
@bronze it's literally an information shredder dude.
complete retards are able to churn out convincing effortposts at light speed. Community IQ is too low to filter these, search engine scoring is too rudimentary to differentiate between "AI post from reddit" and "actual autist gems from the effort mines of deep reddit"
This example I am rageposting about right now is some nigger making claude clone a filter bypass lora for Krea
His post is recent and got upvoted so when you search for the original, the search engines will point you to HIS model on civitai which has a totally incorrect AI-hallucinated explanation.
You know how when you grow up you find more and more things that you just believed as a kid that were just misunderstandings of something, etc? For a while, kid me thought cold turkey meant literally eating turkey cold cuts.
I wonder how much nonsensical babbling people will absorb now that AI auto generates incorrect analogies for every little thing. Feynman talks this issue when interviewer asks him to give one for magnets.
Sometimes you need a simple analogy to advance your knowledge but the more you have and the worse their accuracy... I wonder when things begin to break down. Understanding disappears. Am I making sense?
complete retards are able to churn out convincing effortposts at light speed. Community IQ is too low to filter these, search engine scoring is too rudimentary to differentiate between "AI post from reddit" and "actual autist gems from the effort mines of deep reddit"
This example I am rageposting about right now is some nigger making claude clone a filter bypass lora for Krea
His post is recent and got upvoted so when you search for the original, the search engines will point you to HIS model on civitai which has a totally incorrect AI-hallucinated explanation.
You know how when you grow up you find more and more things that you just believed as a kid that were just misunderstandings of something, etc? For a while, kid me thought cold turkey meant literally eating turkey cold cuts.
I wonder how much nonsensical babbling people will absorb now that AI auto generates incorrect analogies for every little thing. Feynman talks this issue when interviewer asks him to give one for magnets.
Sometimes you need a simple analogy to advance your knowledge but the more you have and the worse their accuracy... I wonder when things begin to break down. Understanding disappears. Am I making sense?
@beardalaxy I keep spinning in circles because I have accidentally ragebaited myself today.
This is a great story btw. It's another issue, the fact that normies are confident enough at all to speak up about anything. Bitch, you don't know computers how the fuck would you know he's savvy
And to add insult to injury, she glazes him but then DOES NOT HIRE him what the fuck dude
This type of babble that normies think is okay to do drives me up the wall. If you say something, then fucking stick with it. I hate when people say shit but don't really mean it (not talking about pleasantries). She obviously takes one stance in conversation but then has another when push comes to shove. This only proves to me that she acts on a whim and that it is totally pointless to reason with her, because she will have forgotten the next day.
AI is kinda similar to this mode of operation because it will just always predict something that "fits" if I'm making sense. It will use the phrase that it thinks is likely next words. "People" that talk this sort of nonsense are the same. They just chain phrases and talking points in hopes it makes them seem smart and to signal they know what they're talking about but they don't. Any convo you have with people like that is effort on your part and prior rote memorization + autopilot on theirs. So is it any wonder she could not tell that his cover letter was AI? It's what she's looking for. "Yes, this phrase was likely to be used here!" so she thinks it's good. Like the AI. Totally hopeless.
Like I said. Wanderer ragebaited himself award.
This is a great story btw. It's another issue, the fact that normies are confident enough at all to speak up about anything. Bitch, you don't know computers how the fuck would you know he's savvy
And to add insult to injury, she glazes him but then DOES NOT HIRE him what the fuck dude
This type of babble that normies think is okay to do drives me up the wall. If you say something, then fucking stick with it. I hate when people say shit but don't really mean it (not talking about pleasantries). She obviously takes one stance in conversation but then has another when push comes to shove. This only proves to me that she acts on a whim and that it is totally pointless to reason with her, because she will have forgotten the next day.
AI is kinda similar to this mode of operation because it will just always predict something that "fits" if I'm making sense. It will use the phrase that it thinks is likely next words. "People" that talk this sort of nonsense are the same. They just chain phrases and talking points in hopes it makes them seem smart and to signal they know what they're talking about but they don't. Any convo you have with people like that is effort on your part and prior rote memorization + autopilot on theirs. So is it any wonder she could not tell that his cover letter was AI? It's what she's looking for. "Yes, this phrase was likely to be used here!" so she thinks it's good. Like the AI. Totally hopeless.
Like I said. Wanderer ragebaited himself award.
@beardalaxy a lot of people just aren't
I try to make my peace with it
sometimes they are just not real because they are sleep deprived / medicated and on autopilot and could possibly put effort into a conversation / thinking under different circumstances. But whether or not a constant or a variable 80% of people at any given time are not "being real" doesn't really matter for the phenomenon.
It happens to everyone at some point that they're just saying shit, but if you're halfway reasonable you try to catch yourself or even simpler just stfu instead of talking nonsense
I try to make my peace with it
sometimes they are just not real because they are sleep deprived / medicated and on autopilot and could possibly put effort into a conversation / thinking under different circumstances. But whether or not a constant or a variable 80% of people at any given time are not "being real" doesn't really matter for the phenomenon.
It happens to everyone at some point that they're just saying shit, but if you're halfway reasonable you try to catch yourself or even simpler just stfu instead of talking nonsense
@WandererUber I used em in college papers and occasionally in effortposts. Basically never in standard shitposts.
@Racist_Columbo I exclusively use Erm-dashes
@WandererUber i dislike that i laughed at that
@bronze @WandererUber it's so bad that even though i occasionally just speak like that i consciously avoid it to avoid giving the impression i'm ai
@All_bonesJones @bronze I think a real actual person with enough awareness to do that is in no danger to consistently sound like AI
@WandererUber @bronze you're absolutely right
@All_bonesJones @bronze @WandererUber Believe none what you hear and half of what you see.
A skeptical mind is a wise one.
A skeptical mind is a wise one.
@Lyle @bronze @WandererUber i will believe whatever makes me feel good and i think is funny regardless of if it's true
@ShovelTrouble >thread about AI using bad analogies
>guy comes in with one that fits perfectly
well well well. If it ain't the indomitable human spirit!
It's one thing to have the models be directly responsible because the output was "cut the blue wire" and that was wrong. I think once this information rot has set in, it will literally be like the Grylls thing. People will have entirely incorrect mental models of the world. This could render them incapable of absorbing new information because of the contradictions and it becomes impossible to tell which part is incorrect.
>guy comes in with one that fits perfectly
well well well. If it ain't the indomitable human spirit!
It's one thing to have the models be directly responsible because the output was "cut the blue wire" and that was wrong. I think once this information rot has set in, it will literally be like the Grylls thing. People will have entirely incorrect mental models of the world. This could render them incapable of absorbing new information because of the contradictions and it becomes impossible to tell which part is incorrect.
@WandererUber It serves as a kind of poison pill for the normal way many people learn things. "I heard someone say this a few times", or "I saw it happen once or twice".
@WandererUber Perfect sense. The models used are geared for people feeling like they understand rather than making sure that they do understand.
It will probably get serious enough to cost lives, like when Bear Grylls took over a more "action" style of survival show when Survivorman was winding down. All those lessons of unnecessary parkour moves and drinking one's own urine worsened the chances of regular people who found themselves in important situations.
@WandererUber yes and no. you're seeing it now in places it was seldom, if ever, used but there's a reason ai adopted its use in the first place. there are contexts where you do see it, like academic writing, so i assume during rlhf the em dash correlated with being seen as erudite, which is where ai gets the habit.
@bronze @WandererUber "it's not just about your airflow, it's about justice"
@HatkeshiatorTND this was more a reply to some people saying the em-dash is not a reliable indicator of something being AI. I know there were places where you'd see it but I also know that it's out of place where I see it now because they didn't used to have it.
Mup da doo didda po mo gub bidda be dat tum muhfugen bix nood cof bin dub ho muhfugga 
@WandererUber entirely correct assessment. there's places it's barely less than a smoking gun and places where it means nothing.
@HatkeshiatorTND didn't we do this whole rigamarole already? I think I've proven that I can distinguish between those very well
@WandererUber That is BS. I used to use them in my old old posts, although I didn’t always have them on my keyboard and sometimes did double dashes.
@WandererUber I don't believe I was taught what an em dash was or how to use it in neither high school nor college. I'm going to assume there was an effort to phase it out.