Twitter as (Limited) Truth Machine

At 7:15pm EST, DJ and former ‘Good Charlotte’ member Josh Madden tweeted a warning to New Yorkers:

WARNING: in Williamsburg people posing as Con Edison workers and knocking on doors DO NOT answer they just robbed someone at gunpoint

— Josh Madden (@Joshmadden) October 31, 2012

I moved to Williamsburg four months ago from Washington, D.C., so this tweet (retweeted by a college friend of mine because, to be frank, I loathed Good Charlotte in high school and could never imagine ever having the desire to get the daily updates of their members) stood out to me. It was unsourced, really, and highly dubious. I had just read John Herrman’s argument at BuzzFeed that Twitter is a truth machine that, under the right circumstances, systematically vets — and inherently destroys — rumors as quickly as it propagates them. “Initial misinformation has consequences, and a consensus correction on Twitter won’t stop any number of these rumors from going viral on Facebook,” he writes. “There, your claims are checked by your friends; on Twitter, if they spread, they’re open to direct scrutiny from people who might actually know the truth.” The argument, I suppose, is that the lightning-fast fact-checking of photos on display during Hurricane Sandy is a formative moment for Twitter and information: if social media can be an echo chamber, truth is louder than fiction.

I decided to try an experiment. I retweeted Madden’s message at around 9:14pm EST. Sure enough, I got two skeptical responses almost immediately in return. The first, at 9:15pm EST, from Digg general manager Jake Levine:
 

@jaredbkeller was reported a while ago, seems false? bit.ly/RrOxWL

— Jake Levine (@jrlevine) October 31, 2012

The second, at 9:24pm EST, from PBS Need to Know web producer Beth Ponsot:

@jaredbkeller Saw that tweet about Williamsburg earlier, but I think it may be of troll origin.

— Beth Ponsot (@bponsot) October 31, 2012

Boom. Lawyered. The Twitter truth machine told me what I needed to know, and I could go back to the Celtics losing to the Heat without worrying about getting pistol-whipped by faux-utility workers on my way home.

But, then. At 10:44pm EST, I got a text message from a number I didn’t recognize. It was obviously a mass text from an iPhone (I have an Android), sent to at least a dozen people: 

Hi Williamsburg/Greenpoint friend. I just read a news alert of two seperate reports of people posing as coned workers, knocking on people’s door and robbing them at gunpoint in williamsburg. I just want to pass along the info. Stay safe and maybe don’t answer your doors…

Two other friends responded with thanks.

This is the problem with the Twitter truth machine: what happens on Twitter doesn’t stay on Twitter. If we’re to continue the favored epidemiology metaphor of the Internet-employed, information that goes viral can become airborne: it leaves the Twitter network, where the journalists and reporters and ‘influentials’ who can quickly propagate corrections don’t have reach. Remember, while a growing portion of the population uses social media as part of their daily news consumption, it’s far from a majority. The latest back of Pew data indicates that social-networking sites grew as a source for news from 9 percent to 19 percent in the last two years; among the social-networking sites, 13 percent of respondents got some news from Twitter, and Twitter users are avid consumers of news. But only 9 percent of the Twitter users said they tweeted or retweeted regularly, which was a similar active user percentage to other social-networking sites.

A fraction of the U.S. uses Twitter, but everyone talks to their siblings, their parents, their coworkers, their friends. Text messaging, email, and ‘dark social’ networks spread misinformation just as fast as Twitter, and to more people. My friends who sent me those messages don’t use, or give any semblance of a shit about Twitter (except sheer astonishment that this is somehow an aspect of my career). They got this information from other friends who read it on Twitter. Even if every single person who retweets a piece of false information (I, for one, fell for the NYSE-under-three-feet-of-water ruse … when CNN reported it) pushes a correction, to what extent is the damage already done?  There are great projects like Truthy and, now, more informal networks that spontaneously snuff out false information, but isn’t it somewhat moot once rumors permeate different networks, ones potentially untouched by that lightning-fast correction? It makes me think or urban legends. And the Loch Ness monster. 

There are perhaps two somewhat practical solutions here: either Twitter integrate some level of correction functionality that propagates a correction to every single connection that’s retweeted, favorited, or, naw, MT’d a false piece of information, or journalists and producers and those with a high Klout appreciation take Cord Jefferson’s advice and remember that, for all it’s benefits as a truthiness crucible, Twitter can also be a dangerous lie machine. It’s much easier to clean up a mess if there’s simply no mess to clean up.    

Update: from Asher Klein at the Chicago Reader:

@jaredbkeller Twitter is part of the new news ecosystem as publisher first, fact checker second. That “truth machine” is a just an apology.

— asher klein (@kleinstar) October 31, 2012

I’m also just going to leave this here: 

I swear to God, if I read one more second-day post-disaster “future of news” story I’m going to put my fist right through my screen.

— Jared Keller (@jaredbkeller) October 30, 2012

Oops.

[Images: Twitter Network Visualizations: “Stanford.” from Analyzing social media networks with NodeXL: Insights from a connected world (Marc_Smith/Flickr)]

11:50 pm  •  30 October 2012  •  9 notes