Alex Stamos, who's spearheading the company's probe into Russian-linked ads placed on its service throughout the 2016 election campaign, on Saturday defended the company's use of algorithms, that verify what users see in their news feeds, parsing out hate speech and threats of violence.
"I am seeing plenty of coverage of our recent problems driven by stereotypes of our workers and attacks against fantasy, strawman school [companies]," he aforementioned during a Twitter thread. "Nobody of substance at the large corporations thinks of algorithms as neutral.
Facebook has drawn flak for its role in perpetuating hoaxes and its influence on the presidential election, and has in recent months worked to combat the increase of faux news. In response, the corporate aforementioned in August it might expand a program to supply connected articles on a trending topic that supply fact-check articles and different views.
The company additionally aforementioned it might use "updated machine learning" to sight a lot of potential hoaxes and send them to third-party truth checkers.
The abundance of faux news on the net within the lead-up to President Donald Trump's election conclusion last year has become a hot-button issue, entangling school giants like Facebook and Google. varied allegations say the pretend news shared on the social networks helped Trump win.
Stamos' team aforementioned in August it had known concerning five hundred "inauthentic accounts" that bought $100,000 value of ads that targeted extremely politicized social problems like immigration, guns and LGBT rights. Facebook has sent records of the ads to government investigators trying into Russia's alleged meddling within the 2016 North American country presidential election.
Facebook did not now reply to asking for comment.
Special Reports: All of CNET's most in-depth options in one straightforward spot.
It's Complicated: this is often qualitative analysis within the age of apps. Having fun yet? These stories get to the center of the matter.
No comments:
Post a Comment