It’s stupidly easy to make up stuff on AITA and get upvotes/comments. I made up one just for fun and was surprised at how popular it got. Well, now not so much, but back when I did.
If you know the audience and what gets them upset, you’ve got easy karma farming.
Two weeks ago someone on one of those story subs, I think it was amioverreacting, was milking off karma making updates. They made 5 posts about the whole thing and even started to sell merch to profit in real life until they took the last post down.
It’s like reality TV & soap operas in text form.
You can somewhat easily spot the AI posts though, which are plentiful now. They all tend to have the same “professional” writing style and a high tendency to add mid sentence “quotes” and em dashes (—) which you need a numpad combo to actually write out manually - a casual write-up would just use the - symbols, if at all. LLMs also make a lot of logic errors that may pop up. Example from one of the currently highly upvoted posts:
He pulled out what looked like a box from a special jewelry store. My heart raced with excitement as I assumed it was a lovely bracelet or a special memento for our wedding day. But when he opened the box, I was absolutely stunned. Inside was a key to a house he supposedly bought for us. I was taken aback because I had no idea he was even looking for real estate. My first reaction was one of shock and confusion, as I thought it was a huge decision that we should have discussed together.
As I processed the moment, I realized the house wasn’t just any house—it was a fixer-upper on the outskirts of town. Now, I get that it can be a great investment, but this particular house needed a ton of work. I’m talking major renovations and repairs, and I honestly had no desire to live there.
Aside from the weird writing (Oh jolly! Expensive gifts! How exciting!), this lady somehow realized & identified this house, location and its state just by looking at some random key in that moment. Bonus frustration if you read through the comments who eat all of this shit up, assuming they aren’t also bots.
Interesting observation about the em dash. I never thought about it that hard, but reddit’s text editor (as well as Lemmy’s, at least on the default UI) automatically concatenate a double dash into an en dash, rather than an em dash.
I use em dashes (well, en dashes, as above) in my writing all the time, because I am a nerd.
For anyone who cares, an en dash is the same width as an N in typical typography, and looks like this: –
An em dash is, to no one’s surprise, the same with as an M. It looks like this: —
(For what it’s worth, Lemmy does not concatenate a triple dash into an em dash. It turns it into a horizontal rule instead.)
I use the poor man’s emdash (two hyphens in a row) here and there as well. I guess I never noticed Reddit auto-formats them. I have been accused of being an AI on a few occasions. I guess this is a contributing factor to why that is.
Funny how Reddit technically formats it into the wrong glyph, though. Not like anyone but the most insufferable of pedants would notice and care, of course. I find it merely mildly amusing.
That’s probably because the posts are stored as plain text, and any markdown within them is just rendered at display time. This is presumably also how you can view any post or comment’s original source. So, here you go:
Double –
En – (alt 0150)
Em — (alt 0151)
And for good measure, a triple:
Actually, I notice if you include a triple that’s not on a line by itself it does render it as an em dash rather than en, like so: —
You’re right, it means you don’t have to save two versions or somehow convert it back into a source format instead.
The triple renders as a line below on mbin. I don’t remember what they’re called.
It’s stupidly easy to make up stuff on AITA and get upvotes/comments. I made up one just for fun and was surprised at how popular it got. Well, now not so much, but back when I did.
If you know the audience and what gets them upset, you’ve got easy karma farming.
Two weeks ago someone on one of those story subs, I think it was amioverreacting, was milking off karma making updates. They made 5 posts about the whole thing and even started to sell merch to profit in real life until they took the last post down.
It’s like reality TV & soap operas in text form. You can somewhat easily spot the AI posts though, which are plentiful now. They all tend to have the same “professional” writing style and a high tendency to add mid sentence “quotes” and em dashes (—) which you need a numpad combo to actually write out manually - a casual write-up would just use the - symbols, if at all. LLMs also make a lot of logic errors that may pop up. Example from one of the currently highly upvoted posts:
Aside from the weird writing (Oh jolly! Expensive gifts! How exciting!), this lady somehow realized & identified this house, location and its state just by looking at some random key in that moment. Bonus frustration if you read through the comments who eat all of this shit up, assuming they aren’t also bots.
Interesting observation about the em dash. I never thought about it that hard, but reddit’s text editor (as well as Lemmy’s, at least on the default UI) automatically concatenate a double dash into an en dash, rather than an em dash.
I use em dashes (well, en dashes, as above) in my writing all the time, because I am a nerd.
For anyone who cares, an en dash is the same width as an N in typical typography, and looks like this: –
An em dash is, to no one’s surprise, the same with as an M. It looks like this: —
(For what it’s worth, Lemmy does not concatenate a triple dash into an em dash. It turns it into a horizontal rule instead.)
I use the poor man’s emdash (two hyphens in a row) here and there as well. I guess I never noticed Reddit auto-formats them. I have been accused of being an AI on a few occasions. I guess this is a contributing factor to why that is.
Funny how Reddit technically formats it into the wrong glyph, though. Not like anyone but the most insufferable of pedants would notice and care, of course. I find it merely mildly amusing.
Does not look like your Lemmy editor is converting them, but Lemmy’s frontend itself, because on mbin I just see two regular dashes. Kind of a weird implementation.
That’s probably because the posts are stored as plain text, and any markdown within them is just rendered at display time. This is presumably also how you can view any post or comment’s original source. So, here you go:
Double –
En – (alt 0150)
Em — (alt 0151)
And for good measure, a triple:
Actually, I notice if you include a triple that’s not on a line by itself it does render it as an em dash rather than en, like so: —
You’re right, it means you don’t have to save two versions or somehow convert it back into a source format instead. The triple renders as a line below on mbin. I don’t remember what they’re called.
That’s a horizontal rule.
Refreshingly, the frontend just converts it to a plain old HTML <hr> tag and doesn’t try to reinvent the wheel, either.
Now that you mention it, I might be the only non-AI using em dashes on the internet (I have a program that joins two hyphens into an em dash).
Apparently Lemmy, and Reddit (I can’t test either one), actually render it that way too. Not sure how many people know about that though.