Meta CEO Mark Zuckerberg announced a series of major changes to the company's moderation policies and practices, saying the election felt like a "cultural tipping point."
Easier to trust and more accurate currently, but I don’t doubt that the algorithm to generate the notes will be internal and closed source, allowing them to utilize that trust to manipulate people.
There is, but doesn’t explain why there’s more upvotes on the post than the comment. Most people would downvote the post after reading that comment, but it’s usually higher anyway. (and sometimes it’s not, I know)
and, what happens when say the community overwhelms, say a conservative facebook group, could add a community note saying “the geese are dissapearing near hatian communities, and there are x missing cats and dogs”. While voting against notes actually reporting the Mayor, Police etc… having denied the claims and also noting that the missing animals are normal for any region of said size.
Nothing stopping meta from adjusting or hiding the vote count later. Hundreds of ways to fiddle with that thing, some really subtle and would not generate drama.
I’m pretty sure the current reason to remove the fact checking is so the company is not put on a collision course with government, a government that will lie so much daily it will break all records.
So, just because they can alter the community notes does not mean they will
Except in instances when the Notes were screenshot and passed around as a joke, I don’t know how many people took them seriously on X, The Everything App.
I don’t know what the “International Fact-Checking Network” is and I doubt most Facebook users do. The type of person using Facebook is going to likely trust notes written by their peers more than things that come from “on high” (meaning Facebook themselves)
I suppose I’m just seeing how even Twitter has had success with community notes, and figured it would be the same on Facebook. But it’s easy to forget just how… out there Facebook is these days.
Community Notes are good, but they’re never a complete replacement for paid work. And my second paragraph is based on some notable incidents on X; it’s not just “oh it’s only bad because it’s on Facebook”.
Honestly not a bad idea. The community notes are easier to trust and typically more accurate anyway.
Easier to trust and more accurate currently, but I don’t doubt that the algorithm to generate the notes will be internal and closed source, allowing them to utilize that trust to manipulate people.
Community notes are written and voted on by the community
Because that’s never gone wrong before
I trust that more than some random fact checkers tbh.
Just look at front page on reddit. Basically half of the headlines are misleading.
A lot of the time there’s a comment correcting the title or article at the top though. Pros and cons with that system.
There is, but doesn’t explain why there’s more upvotes on the post than the comment. Most people would downvote the post after reading that comment, but it’s usually higher anyway. (and sometimes it’s not, I know)
Because people don’t go to the comments, they read the title on the front/subreddit page, sometimes vote and then move on.
Exactly my point in my comment above the other.
Yes, and reddit has massive centralized ban squads that suppress dissent.
and, what happens when say the community overwhelms, say a conservative facebook group, could add a community note saying “the geese are dissapearing near hatian communities, and there are x missing cats and dogs”. While voting against notes actually reporting the Mayor, Police etc… having denied the claims and also noting that the missing animals are normal for any region of said size.
Nothing stopping meta from adjusting or hiding the vote count later. Hundreds of ways to fiddle with that thing, some really subtle and would not generate drama.
I’m pretty sure the current reason to remove the fact checking is so the company is not put on a collision course with government, a government that will lie so much daily it will break all records.
So, just because they can alter the community notes does not mean they will
In a capitalist society, you get much better quality when you pay someone their living to do that.
Except in instances when the Notes were screenshot and passed around as a joke, I don’t know how many people took them seriously on X, The Everything App.
Not when the community notes will be written by AI, and voted on by bots.
Whomever has the most AI and bots to swamp the notes with their text and generate votes wins.
Does that sound like a good way to get facts?
It’s a great way to get facts if you want your facts to be accurate less than half of the time.
Better than partners certified by the International Fact-Checking Network?
And who certified the international fact checking network!?
International Board of Fact-Checking
And who fact checks the International Board of Fact-Checking?!
They are periodically reviewed by the International Congress of Fact Checking Board Auditors
I don’t know what the “International Fact-Checking Network” is and I doubt most Facebook users do. The type of person using Facebook is going to likely trust notes written by their peers more than things that come from “on high” (meaning Facebook themselves)
just search it up gal
How is that a good thing if a lot of these notes take content out of context or are just plain wrong, echoed by those who trust misinformation?
I suppose I’m just seeing how even Twitter has had success with community notes, and figured it would be the same on Facebook. But it’s easy to forget just how… out there Facebook is these days.
Community Notes are good, but they’re never a complete replacement for paid work. And my second paragraph is based on some notable incidents on X; it’s not just “oh it’s only bad because it’s on Facebook”.