Davey AlbaDec 19, 2022

NEW: Proud that the pub I work for is doing some of the best analysis of Twitter's features promoted under the Musk era.

Here's a new analysis of the fact-checking darling Community Notes, which *DOESN'T* display context notes on Twitter's most divisive tweets https://www.bloomberg.com/graphics/2022-twitter-birdwatch-community-notes-misinformation-politics/

Show threadDavey AlbaDec 19, 2022
“Far fewer users are classified as being in the middle. In order for notes to be publicly displayed, the algorithm requires ‘helpful’ votes from users on both sides of the political spectrum.” https://www.bloomberg.com/graphics/2022-twitter-birdwatch-community-notes-misinformation-politics/
Show threadDavey Alba
"Abortion is never medically necessary to save the life of a mother," reads a tweet from
@[email protected]. This is factually untrue, but because more “left-leaning” accounts voted on the contextual note, the note is suppressed & Twitter says it “needs more ratings.” https://www.bloomberg.com/graphics/2022-twitter-birdwatch-community-notes-misinformation-politics/
Dec 19, 2022 at 1:46pmWeb
Show threadpgcdDec 19, 2022
@daveyalba So Reuters and AP are left-leaning? Cool.
Show threadDavey AlbaDec 19, 2022
@pgcd If you believe the way Twitter users are "assigned" to this left-right spectrum.... then I think that's how Community Notes would also assign Reuters and AP.
Show threadkeplernikoDec 19, 2022
@daveyalba wait so my vote is worth less (same as others like me) because *I have a particular left/right bias* *as determined by an algorithm*? This would seem to run afoul of data protection laws like GDPR and others.
Show threadDavey AlbaDec 19, 2022
@keplerniko Yepppp
Show threadWetHat💦Dec 19, 2022
@keplerniko @daveyalba That might also be a violation of civil rights (the rights of citizens to political and social freedom and equality.)
Show threadFreebirdFrolicDec 19, 2022
@daveyalba you will only find shite.
Show threadPeter PrincipleDec 19, 2022
@daveyalba Space Karen is giving Lies Social a real run for its money.
Show threadDavid RolfeDec 19, 2022
@daveyalba seems like a pretty significant flaw in their meta-moderation if it fails on politicized facts.
Show threadDavey AlbaDec 19, 2022
@Dorolfe you would think, right?
Show threadDrewDec 19, 2022
@daveyalba @Dorolfe what a disingenuous metric. Left of what? How do measure each person's "leftness"?
Show threadDavid RolfeDec 19, 2022
@Drew whether or not they follow proud boys or trump — haha it’s probably pretty easy and interesting to graph the network by political follows.
Show threadDavid RolfeDec 19, 2022
@Drew dunno if Twitter employed political science researchers or subject area experts for this stuff. They could’ve?
Show threadDrewDec 19, 2022
@Dorolfe but just because you follow someone, doesn't mean you agree with them
Show threadDavid RolfeDec 19, 2022
@Drew oh I agree! There’s absolutely complexities in interpreting that graph, but with millions of accounts you can probably suss out policy preferences by tweet analysts, rt behavior, along with the raw social graph etc.
Show threadDavid RolfeDec 19, 2022
@daveyalba lolsob this gives me we’re all doomed vibes.
Show threadGreggHillPACDec 19, 2022
@daveyalba suppression of free speech at every corner.
Show threadgormanDec 19, 2022

@daveyalba

Disgusting is the right word @GreenFire
Idiotic too, because in science there should not be "left and right", only "right and wrong".

Show threadMichael KnepprathDec 19, 2022
@daveyalba I feel like there may be a major unsaid issue here. The note a user is rating may be on the same topic that the algorithm used to categorize them left/right to begin with. It’s circular, no?
Show threadcuddly but deadly! Dec 19, 2022
@daveyalba That is amazing. Hard to even be mad because the brazenness of it is just so shocking.