A consequence of covering the intersection of social media and democracy is that sometimes you wind up having to discuss things that are very dumb. The somewhat infuriating controversy over Twitter’s “shadow banning” of prominent conservatives — something that it is in no way doing — is one of them. And yet how Twitter reacts to the attendant criticism could determine whether the company ever gets a handle on the abuse its platform is so well known for.
Yesterday I mentioned a misleading story in Vice whose headline then stated, falsely, that “Twitter is ‘shadow banning’ prominent Republicans like the RNC chair and Trump Jr.’s spokesman.” (It has since been changed.) That story drew from a Sunday piece by Gizmodo that described how some “controversial” accounts were being “demoted in search results.”
The first thing to note about this story is that it begins and ends with which accounts are suggested when one begins typing in a name in the Twitter search box. That’s it. The very worst thing Twitter can be accused of here is, in some cases, making you spell out some people’s full names if you wanted to read their tweets.
Like I said: very dumb.
This story starts last year, when Twitter belatedly began to remove low-quality and harassing tweets from search results. Twitter continued to refine search results based on many different signals, most of which it has not and will never share with us. One effect of this was that the company sometimes would not automatically suggest certain accounts to you, based on the behavior of that account and the accounts that interacted with it.
The operating theory here is that you can tell a lot about a Twitter account from its friends, and if it hangs out with filth, it might not deserve a place in your personalized autocomplete results. Twitter, of course, has been under enormous pressure to make changes like these, following a decade in which it struggled to get abuse under control.
Wired’s Issie Lapowsky talked to Twitter about what might have happened to demote a handful of conservatives in search results:
Twitter has been far from transparent in defining that bad behavior, but a few examples it’s given publicly include accounts that haven’t confirmed their email addresses or that signed up several accounts at once. Twitter doesn’t ban these accounts or the tweets they post. It instead reduces their visibility in users’ replies and also in search. The company’s algorithms also analyze who those accounts are connected to and whether accounts in those networks are also exhibiting troll-like behavior. But Twitter insists the algorithms have no way of knowing whether the people behind those tweets are Republicans or Democrats.
“If you send a tweet and 45 accounts we think are really trolly are all replying a hundred times, and you’re retweeting a hundred of them, we’re not looking at that and saying, ‘This is a political viewpoint.’ We’re looking at the behavior surrounding the tweet,” the spokesperson said.
Meanwhile, President Donald Trump was rage-tweeting about the Vice story. “Twitter ‘SHADOW BANNING’ prominent Republicans,” he said. “Not good. We will look into this discriminatory and illegal practice at once! Many complaints.” (For the record, it is not illegal to make someone spell out a person’s complete name to find their account.)
One negative consequence of today’s news cycle is to strip the word “shadowban” of its meaning — another once-useful piece of platform jargon, like “fake news,” that now barely coheres as an idea. As Brian Feldman recounts in New York, a shadowban — which allows a user to continue posting on a forum, without informing them that their posts are now visible only to them — “is useful in that rather than someone immediately being locked out and possibly retaliating by, for example, making a new account, the user fades away gradually due to the lack of interaction from other users.”
Thanks to the events of this week, though, “shadowban” seems fated to mean “getting less distribution than I personally think it should.” That was the idea that motivated this year’s disastrous Diamond and Silk hearing in Congress, and Twitter’s search hiccup could lead to yet another one.
But let’s be clear: the argument that Twitter has systematically disadvantaged conservative voices can only be made in bad faith. Just as the argument that Facebook systematically disadvantages conservative voices can only be made in bad faith.
Just because it was made in bad faith, however, doesn’t mean it can’t be effective. That was one of the lessons of Facebook’s last two years, in which a Gizmodo story that argued Facebook was “suppressing conservative news” led to the company eliminating human editors from its platform. That helped pave the way for the spread of misinformation on the platform that continues today.
And that’s why a very dumb story about Twitter search results is more consequential that it might first appear. Platforms that find themselves at odds with the president of the United States will be highly prone to overreaction — in ways that make the platform worse. The work Twitter has undertaken to reduce abuse has been welcome. And if it means typing a few more characters into a search box, I say so be it.
“We do not shadow ban,” Twitter said in a statement late Thursday night. “You are always able to see the tweets from accounts you follow (although you may have to do more work to find them, like go directly to their profile). And we certainly don’t shadow ban based on political viewpoints or ideology.”
Update, 11:40 a.m.: This article has been updated to include Twitter’s statement.
Feldman’s whole explainer about the history of shadowbanning here is useful, and I suggest reading the whole thing.
Here are some new stats from Pew on how Americans get their news from social platforms:
Since 2013, at least half of Twitter users have reported getting news on the site, but in 2017, with a president who frequentlymakes announcements on the platform, that share has increased to about three-quarters (74%), up 15 percentage points from last year. On YouTube, about a third of users now get news there (32%), up from 21% in 2016. And news use among Snapchat’s user base increased 12 percentage points to 29% in August 2017, up from 17% in early 2016.
Facebook’s collapse in after-hours trading yesterday continued on Thursday, to record effect:
After a surprisingly weak growth forecast in this week’s earnings report, Facebook’s stock price dropped 19 percent today. The decline, which erased about $120 billion in market value, is the largest one-day drop in the history of the American stock market.
Here’s a fun Deepa Seetharaman that goes minute by minute through the conference call resulting in Facebook’s record stock plunge. Here’s something that hasn’t gotten enough attention yet. (It will though.)
He partly blamed “currency headwinds” and new privacy options for users but also revealed that new ad formats such as those within Instagram Stories weren’t pulling in the same amount of money as ads shown in the Facebook and Instagram feeds.
This revelation about Stories startled many analysts and investors. Facebook executives have said users were embracing the Stories feature, which allows users to post photo and video montages that disappear after 24 hours, and that activity would eclipse time spent just scrolling through feeds next year. Ads shown in feeds are where Facebook generates the bulk of its revenue. Now, Facebook executives were saying that people were spending more time using a less-lucrative product.
Here’s a good nugget from Fred Vogelstein on Facebook’s earnings:
How much less profitable will Facebook be? During the last quarter of 2017, the ratio of operating earnings to revenue—an important measure of profitability—was 57 percent. It was 44 percent in the second quarter of 2018. And it’s expected to fall into the mid-30 percent range by the end of this year, said David Wehner, Facebook’s chief financial officer.
Facebook is looking to beef up Workplace with some run-of-the-mill collaboration tools. Redkix is “an email startup that combines email, messaging and calendar features into one app,” Kurt Wagner reports.
Slack also bought a Slack competitor — Hipchat, which predated it, and which Slack utterly routed it, to the point that Atlassian paid its rival (in the form of an investment) to be rid of it.
“Flop accounts” on Instagram are where teens are discussing real-world events, because they no longer trust the mainstream media, reports Taylor Lorenz:
The accounts post photos, videos, or screenshots of articles, memes, things, or people considered a “flop,” or, essentially, a fail. A flop could be a famous YouTuber saying something racist, someone being rude or awful in person, a homophobic comment, or anything that the teen who posted it deems wrong or unacceptable. Sometimes the teens who run a given account know each other in real life; more likely, they met online.
“Flop accounts bring attention to bad things or bad people that people should be aware of. We also post cringeworthy content for entertainment purposes,” said Alma, a 13-year-old admin on the flop account @nonstopflops.
Here’s a memorably disturbing tale from Kashmir Hill about a woman who was defamed by a disturbed meth addict after making a comment online about a teen’s right to take a selfie at Auschwitz (?????).
I’m sorry but this is the funniest thing I have ever read. Never change, LinkedIn. Unless you’re adding voicemail in 2018 in which case do exactly that.
Have you ever typed out a long message and thought about how much faster and easier it would be to say it out loud? To give you more ways to have conversations, we’ve now added the ability to record and send voice messages up to one minute in LinkedIn Messaging.
Snap has belatedly introduced a way for influencers to make money on Snapchat.
Are you completely ignoring IGTV? No worries, it will now be inserted into your Instagram feed as an enticement to click.
Matt Levine is wonderfully droll on Facebook’s record stock decline:
There is a popular, slightly tongue-in-cheek notion in the financial industry that losing a billion dollars of client money is a badge of honor, something to be proud of, a good thing to have on your resume. It shows you can bounce back from adversity, etc., but more importantly it shows that clients trusted you with a billion dollars and you had the confidence to take risks with it. Losing $150 billion of shareholder money shows, at a minimum, that Facebook had shareholders who believed in it to the tune of almost $630 billion (as of yesterday’s close); you can’t lose $150 billion of market cap without first having $150 billion of market cap. I submit that losing $150 billion of market cap in a day is a more impressive financial accomplishment than anything that almost any other company has done in the history of stock markets.
And finally ...
The hottest trend among teens is spamming Twitter with the least-used emoji so that it is … no longer the least-used emoji. My colleague Shoshana Wodinsky reports:
During the latter bit of those eleven weeks, the tramway became a bit of a hero among a number of public transportation advocates. The campaign to save the tramway emoji from its ignominious fate really took off when the 100,000 members of New Urbanist Memes for Transit-Oriented Teens — a Facebook group dedicated to transportation memes — caught wind of the situation. Dozens of NUMTOTs, as the members call themselves, spammed Twitter with strings of the lone gondola to try to bring it up from its spot in last place.
God bless NUMTOTs and thank you for all that you do.
Talk to me
Questions? Comments? Shadowbans? email@example.com