They don’t have to, algorithms do whatever they are designed to do. Long division is an algorithm.
Profit motives are the issue here.
Algorithms [based on engagement]
Yeah, the narrowing of the word “algorithm” to only mean “social media recommendation algorithms” is getting on my nerves.
It’s the only time normies encounter the word.
How do you feel about crypto?
Cryptography is pretty useful
Scam from start to finish
At this point, “crypto” is its own word short for “cryptocurrency”, and not for “cryptography” in the broader sense. It’s unfortunate, but that’s how people use it.
That isn’t even generally true. Try mentioning crypto on the LKML and see what they think you mean.
I’ve always thought it was so funny when people say tHe aLgOrItHM like it’s a bad word or something. I know they mean social media & marketing, but it’s funny to think that they’re very concerned about something like bubble sort.
They did a study around the 2020 elections and have found the following to work with trolls:
Respond once with the facts (if you must), and then walk away. I have found Lemmy not needing that most of the time, just downvoting seems to work. But if you’re on the place that shall not be named, this works.
I wish lemmy had the feature where you can mute all replies to a comment.
Yes, that and the tagging of users.
Sorry I see you replied before I edited my comment. Don’t feel forced as it isn’t that important but if you have the original study handy I’d appreciate a link because it sounds interesting.
Also you can tag users.
Thanks, I didn’t realize that worked here. But I meant tagging a user as awesome or a troll or whatever. That way, when you kind of remember seeing the name and they seem like they’re trolling, I can tell right away if I’ve had previous interactions with them. RES was awesome for that.
With Boost for Lemmy you can tag people atleast.
Ah that makes sense. That would be nifty. Because my block list these days is based on “this person said fucked up shit and i vaguely remember them saying more fucked up shit a while ago but cant fully remember if its them”
Exaxtly that, I just don’t have the bandwidth.
What I don’t get about this is why in this day and age with all the analytics tools we have do companies continue to just happily pay for simple eyeball exposure?
The only time they seem to have any pause at all on this model is if people post screenshots of ads for their products next to posts literally praising Nazis.
These so called AIs (LLMs) can learn to tell the difference between positive/happy/uplifting posts, neutral posts, and angry/sad/disturbing posts. The advertisers should be asking for their products to be featured next to the first and second groups of posts.
People engage based on anger, sure. They click posts and reply and whatnot. But do they click the ad next to a post that pisses them off and then buy the product?
Or is this purely a subconscious intrusion effort? Do the advertisers just want their products in front of eyeballs regardless of what’s around the ad? It seems like the answer is “no” when they’re called out. But maybe it’s “yes” if they can get away with it?
Is that cat from poland?
no he’s from bangladesh
But… I love this cat and want my fingers to engage with their fuzzy head!
I think the word is ragebait
Every social media company’s content algorithm should be open source or at least a government agency should have code enforcement
I’ve been participating in Threads (yeah, I know, should be ashamed) and I’m unfortunately a sucker for some of the ragebait, especially political.
Guess what Threads pushes at me. A lot of the dumbest ragebait. Not people that actually want to have a conversation. My fault for being a sucker, but the algorithms work.
Doesn’t really matter, I’m shadowbanned. Pissed off too many republican propagandists by refuting them, so as usual, the “report” button is their remedy.
Algorithms simply determine which posts will get the most interaction and feed it to people. Does it benefit corps? Of course! But it’s driven by people who choose to engage in this content.
deleted by creator
Wasn’t this literally the shady research that Facebook got caught doing with Cambridge Analytica? Specifically tweaking a user’s feed to be more negative resulted in that user posting more negative things themselves and more engagement overall.
Yep!
Facebook figured out how to monetize trolling.
Over 10 years later, it’s destroyed society, but made them a lot of money.
I wonder exactly how much of Hawaii Zuckerberg has to own before people start to question what they are getting from facebook.
whose cat is this
Why? You hungry?
…as opposed to platforms like Lemmy, where the only political ideologies you’ll find are “leftists” who, when asked what they even believe, respond with “what are you, a cop?”
are you? 👮♂️ 👀
If you’re a cop, you have to tell me, man! Like legally, you can’t arrest me without telling me you’re a cop!
Nobody likes Muhammad ibn Musa al-Khwarizmi anymore, I guess he had a good run since around 820 CE.
For a long time Facebook counted an angry react as equal to five likes for measuring engagement. It’s very much intentional.
