Facebook Users Cry 'Censorship' After Being Told Which Russian Troll Pages They Liked (gizmodo.com)

An anonymous reader quotes a report from Gizmodo: As the FBI's investigation into Russian election interference reaches a fever pitch, Facebook rolled out a new News Feed alert Monday night. The bulletin told users who followed pages created by Russian trolls that those pages have been removed. And some of the affected users did not like this. A brief search revealed that numerous people believe that this is an act of censorship by Facebook. Some users argued that they should be allowed to decide what's "true, fake, or otherwise," a challenge that's bound to be a slippery slope in this era of algorithm-based confirmation bias. Others took on a more conspiratorial tone, claiming that Facebook failed to reveal which pages were removed (despite the alert containing a link listing the pages in question). Facebook first released the information in December, creating a help page that showed users if they liked or followed pages and accounts associated with the Internet Research Agency, Russia's notorious troll farm, but today's alert seems to have inspired newfound alarm. The fact that Facebook explicitly stated which pages were deleted seems to have done little to reduce the anger over the allegedly clandestine silencing.

Comments

Re:why fb users are dumb

[Lab Rat Jason]

How many of the notified/complaining users are in fact MORE Russian misinformation accounts that have not yet been discovered... and the outcry is simply a ploy to destabilize Facebook?

Re:Denial

[hey!]

I don't think it was infighting, I think it was complacency and misreading the data.

Clinton's campaign acted all along as if they could leave it to Trump lose the election, and the polls seemed to be bearing that out. However polling figures aren't as reliable as the "margin of error" figures suggest, because that margin only represents random sampling errors. It does not account for systematic sampling errors.

Every poll is adjusted by some kind of likely voter turnout model, and in state after state anomalously high rural turnout knocked those models into a cocked hat. The thing is there were warning signs of this from Clinton's own campaigns in those states, which Clinton chose to ignore because the numbers were telling her what she expected to hear.

That's always a danger when you manage by numbers. Numeric and anecdotal data both have their place, mostly to raise healthy doubts about the other.