Nieman Foundation at Harvard
HOME
          
LATEST STORY
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
ABOUT                    SUBSCRIBE
June 4, 2018, 11:40 a.m.
Audience & Social
LINK: newsroom.fb.com  ➚   |   Posted by: Shan Wang   |   June 4, 2018

Against all odds, Facebook’s Trending Topics lasted more than four years.

The company debuted a mostly algorithmic module that pulled in stories posted across the network, based on users’ interests and posts that were getting a lot of increasing attention, in early 2014.

This month, Alex Hardiman, Facebook’s head of news products , announced in a blog post that the company will remove the module sometime this week. Hardiman said the box was only available in five countries anyway and drove less than 1.5 percent of the clicks from the platform to news publishers’ sites, “on average.” Facebook is also shutting down any third-party integrations that have been relying on the Trending API.

For a couple years, the types of stories surfaced in Trending Topics were subject to some mockery (how did “Deez Nuts” come to be a “Trending Topic”?) and criticism. But the heaviest scrutiny came during the height of the 2016 U.S. presidential campaign, when former Facebook contractors who provided some editorial oversight to the Trending Topics module claimed to Gizmodo that other “curators” were suppressing conservative stories and sources. (The initial Gizmodo story had called Trending Topics among “the most powerful real estate on the internet.”)

What followed was, as Wired called it, Facebook’s two years of hell, as the company reckoned with fake news, abuse of its platform, and increasingly loud conservative voices crying foul. For Trending Topics, this reckoning first meant losing its human oversight — Facebook laid off all the contracted “news curators” a couple of months after the first Gizmodo story. Its ultimate shutdown this week has been a long time coming.

Facebook is now instead touting three other new initiatives it’s been testing with news publishers, noting in particular that it’s seen that “the way people consume news on Facebook is changing to be primarily on mobile and increasingly through news video”:

— A breaking news label for the News Feed it’s testing with 80 publishers across North America, South America, Europe, India and Australia, as well as breaking news notifications.

— A Today In section focused on local news and community goings-on, being tested in a few dozen cities. Some news organizations based in test cities whose stories sometimes get pulled into the Today In module have told me the stories there are sometimes more than a day or two old, and generate no meaningful traffic to their sites.

— News video coming to Facebook’s dedicated “Watch” section in the U.S., where “people can view live coverage, daily news briefings and weekly deep dives that are exclusive to Watch.” These shows are close to launching, according to The Wall Street Journal, and will “likely include content from Fox News and CNN” and will be funded by Facebook and run exclusively on the Facebook.

Reality check #1: Babies and puppies videos topped this one analysis of the most engaged native videos on Facebook through May 2018. Among traditional news publishers, ABC News and Fox News each had one video make it to the 100 most engaged Facebook videos. Reality check #2: Facebook paid millions to publishers and celebrities to create a certain amount of live video for the platform; when Facebook stopped paying, the major publishers quickly scaled down their productions.

Show tags
 
Join the 60,000 who get the freshest future-of-journalism news in our daily email.
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
Within days of visiting the pages — and without commenting on, liking, or following any of the material — Facebook’s algorithm recommended reams of other AI-generated content.
What journalists and independent creators can learn from each other
“The question is not about the topics but how you approach the topics.”
Deepfake detection improves when using algorithms that are more aware of demographic diversity
“Our research addresses deepfake detection algorithms’ fairness, rather than just attempting to balance the data. It offers a new approach to algorithm design that considers demographic fairness as a core aspect.”