Tonight, in the aftermath of yet more mass shootings, the internet is once again a swirl of threads. Debates about gun control, white supremacy, domestic terrorism, video games, mental illness, de-platforming, content moderation, and regulation all rage simultaneously on cable news, in print, and on Twitter. The mass shootings in El Paso and Dayton — like so many mass shootings before them — overwhelm us in their horror. But we owe a debt to the victims to sort through it as best as we can, and to find a way forward.
Around here we’re primarily interested in the relationship between tech platforms and our democracy. And yet while much attention has focused on the role that 8chan and other hate sites played in fomenting these attacks, the more I worry that the attention is misplaced. We have already identified very good solutions to the problem of mass shootings — and while we can and should talk about what to do with hate platforms, we shouldn’t lose sight of two much more important factors in the past decade’s spike in mass shootings.
The first and most important factor in this weekend’s shootings is, of course, the easy availability of guns. In 2017, Max Fisher and Josh Keller wrote that the United States had 270 million guns and had 90 mass shooters between 1966 and 2012. No other country had more than than 46 million guns or 18 mass shooters.
In 1994, the United States banned assault weapons, and mass shootings dropped by 43 percent. Republicans allowed the ban to expire a decade later, and mass shootings increased by more than 230 percent. At least, those are the estimates that I found today. We don’t even really know how many people are shot each year in the United States, thanks to aggressive lobbying by the gun industry.
The Dayton shooter apparently had no clear racial or political animus, according to CNN. But he did have access to an AR-15-style rifle and 100-round drum magazines.
The relationship between social networks and democracy is dynamic and unpredictable. The relationships between guns and mass shootings is not. The more guns that are on the streets, the more people die, and any discussion of platform culpability should come after gun control.
The second crucial factor in the spate of mass shootings is the United States’ long and awful history of white supremacy. The El Paso shooter’s manifesto protesting a “Hispanic invasion” closely resembles language used ad nauseam by the president and Fox News. As CJ Werleman writes in the Sydney Morning Herald:
It was only two weeks ago when Trump inspired an auditorium full of his supporters to chant “send her back” in reference to the country’s first elected black Muslim congresswoman, Ilhan Omar, who was born in Somalia and migrated to the US as the young daughter of refugee parents.
Earlier in the year, Trump smeared all immigrants approaching the US-Mexico border as invaders when he said, “People hate the word ‘invasion’, but that’s what it is.”
The concerted effort among Republican politicians and conservative media outlets to promote the idea of an “invasion” has given cover to white terrorists who want to commit mass murder. We talked about this “stochastic terrorism” last October:
I encountered the idea in a Friday thread from data scientist Emily Gorcenski, who used it to tie together four recent attacks.
In her thread, Gorcenski argues that various right-wing conspiracy theories and frauds, amplified both through mainstream and social media, have resulted in a growing number of cases where men snap and commit violence. “Right-wing media is a gradient pushing rightwards, toward violence and oppression,” she wrote. “One of the symptoms of this is that you are basically guaranteed to generate random terrorists. Like popcorn kernels popping.”
It was clear in October and remains so today: so long as the president, Fox News, and others promote the idea that immigrants and nonwhite citizens are destroying the country, more popcorn kernels will pop — and we’ll see more of the rampages we saw this weekend in El Paso. (It seems fitting that another rampager, Cesar Sayoc, was sentenced to 20 years in prison today for mailing bombs to the president’s political nemeses.)
The combination of easily available guns and the ruling party promoting a sense of perma-crisis is so powerful that I believe we would see mass murders even if 8chan had never existed. But much effort has been expended over the weekend in unpacking the role of hate sites in these shootings, so let’s see what we can learn.
8chan, which was founded because its users posted content so vile that even the free-speech bastion 4chan wouldn’t tolerate it, became a focus of attention this weekend when the El Paso shooter posted his manifesto there. He was the third shooter to post a manifesto on 8chan before committing murder in the past six months.
Robert Evans, who has studied 8chan extensively, says 8chan contributes to domestic terror by (1) creating a welcome home for people to discuss and promote their crimes and (2) “gamifying” terrorism by encouraging people to beat the previous shooter’s “high score” — murdering more people than the previous shooter. Evans concludes:
In the wake of the Christchurch shooting I published my first Bellingcat article about 8chan. I was interviewed by numerous media agencies about the website, and I warned all of them that additional attacks would follow – every month or two – until something was done. This prediction has proven accurate. Until law enforcement, and the media, treat these shooters as part of a terrorist movement no less organized, or deadly, than ISIS or Al Qaeda, the violence will continue.
“Shut the site down,” Mr. Brennan said in an interview. “It’s not doing the world any good. It’s a complete negative to everybody except the users that are there. And you know what? It’s a negative to them, too. They just don’t realize it.”
On Monday, 8chan went down after web services provider Cloudflare ended its relationship with the site. Everyone presumes it will find some sort of service provider eventually, though it has struggled to find new service providers all day. Much of the discussion today focused on what practical effect shutting down 8chan might have on future terrorist attacks. I’m skeptical it will have much, though I’m glad 8chan is on the ropes — eliminating sites that host terrorist content makes it harder and more expensive for future terrorists to set up shop and recruit others to their cause.
President Trump gave a speech in which he suggested social platforms should become adjuncts of law enforcement, working with the Department of Justice “to develop tools that can detect mass shooters before they strike.” It it not clear how such a tool would work, and having Facebook or Twitter analyze every post and referring suspicious ones to the government would seem to infringe greatly on our civil liberties.
Notably, the government was building tools to counter violent extremism in 2016, before Trump pulled the funding. As Peter Beinart recounted in The Atlantic last year:
In the waning days of Barack Obama’s administration, the Department of Homeland Security awarded a set of grants to organizations working to counter violent extremism, including among white supremacists. One of the grantees was Life After Hate, which The Hill has called “one of the only programs in the U.S. devoted to helping people leave neo-Nazi and other white supremacy groups.” Another grant went to researchers at the University of North Carolina who were helping young people develop media campaigns aimed at preventing their peers from embracing white supremacy and other violent ideologies. But soon after Trump took office, his administration canceled both of these grants. In its first budget, it requested no funding for any grants in this field.
It’s part of a pattern of neglect. The grants were administered by the Office of Community Partnerships, which works intimately with local governments and community organizations to prevent jihadist and white-nationalist radicalization. In Obama’s last year, according to the former director, George Selim, the office boasted 16 full-time employees, roughly 25 contractors, and a budget of more than $21 million. The Trump administration has renamed it the Office of Terrorism Prevention Partnerships, and cut its staff to eight full-time employees and its budget to less than $3 million.
It’s little wonder then, that in the wake of the attack, six former counterterrorism directors for the National Security Council wrote a letter calling for “a significant infusion of resources to support federal, state, and local programs aimed at preventing extremism and targeted violence.”
Programs that work to reduce extremism can be useful. So can disrupting internet forums where terrorists organize and support one another. But it would be a mistake to focus on how easy it is for terrorists to meet on the internet rather than on how easy it is for them to buy AR-15s. We can’t predict the effect of shutting down 8chan and forcing their users to disperse. But we can predict the effect of taking automatic weapons out of their hands.
I pray that we will, and soon.
More on the shootings
A roundup of platform policies on white supremacy. A roundup of hoaxes about the shootings. The 151 companies that provide web services to hate groups. The 2,200 times Trump ran a Facebook ad with the word “invasion” in it. Six of the 10 deadliest acts by domestic terrorists were in the past five years. Why “there are no lone wolves” among mass shooters.
And finally, the Ohio state lawmaker who placed blame for the shootings on, among other things, “drag queen advocates.”
John D. McKinnon and James V. Grimaldi report that there’s a conflict between the federal agencies taking a lead on antitrust cases:
The Justice Department’s antitrust division has flexed its muscles in ways that have rankled staff at the FTC, most particularly in an antitrust lawsuit against Qualcomm Inc.After the FTC prevailed in a 10-day trial in federal court in San Francisco in May, FTC lawyers were taken aback when Justice Department lawyers sided with Qualcomm in the appeal phase of the litigation.
The two agencies, which have divvied up antitrust enforcement for more than a century, have been known to skirmish over their shared responsibilities in the past. But some observers say it has reached new heights just as the scrutiny of companies such as Alphabet Inc. ’s Google unit, Amazon.com Inc., Facebook Inc. and Apple Inc. has become a priority in Washington.
The bad bots are back, Maureen Linke and Eliza Collins report:
Hundreds of social-media accounts with bot-like traits promoted misinformation and content aimed at inflaming racial divisions during both nights of Democratic presidential debates, continuing similar activity during the first set of debates in June, according to data analyzed by The Wall Street Journal.
The bot-like activity on Tuesday and Wednesday nights was consistent with online discussions around Sen. Kamala Harris’s ethnicity during the first Democratic debate, indicating both Ms. Harris, who is black, and racial issues as key targets for bot-like accounts during the 2020 election campaign.
Thomas Brewster reports that Microsoft is funding technology that is being used to build surveillance systems in Palestine:
Microsoft is under fire for funding the Israeli facial recognition company AnyVision, which is reportedly carrying out surveillance on Palestinians. AnyVision also supplies technology in Russia and Hong Kong, where human rights are under attack.
Privacy activists say it’s another sign Microsoft is pushing the controversial technology, despite presenting itself as more progressive and transparent on the ethics of facial recognition than rivals like Amazon and Google. Shankar Narayan, the director of the Technology and Liberty Project at the American Civil Liberties Union (ACLU), told Forbes that he’d held meetings with Microsoft in Seattle last year in which the tech giant appeared receptive to ideas on holding back the spread of facial recognition. But the company has not followed through with any action, Narayan claimed.
Lauren Feiner reports on a novel approach to avoiding Twitter abuse:
Twitter users often point to the company’s content policy in Germany to argue it should be able to identify and remove Nazis from the platform in other regions. When Maureen Colford learned about the location setting “hack” to filter out Nazis, she said she was “amazed that somehow Twitter manages to do this in Germany,” and wondered, “why can’t they do this everywhere?”
This theory assumes Twitter has a filter it uses to detect hate speech in Germany and chooses not to implement it elsewhere. But that’s not how it works. Twitter is simply required by German law to remove some forms of hateful content expeditiously.
Tech companies have reported back to Congress on how they plan to address deepfakes, Rebecca Kern reports:
Facebook, Twitter and Google are considering policy changes on handling realistic but fake videos and images following a widely circulated doctored video of Speaker Nancy Pelosi, the companies told a key lawmaker.
Rep. Adam Schiff (D-Calif.) asked the three companies in July 15 letters to detail how they planned to address doctored videos and images, including “deepfakes,” which are manipulated content created with artificial intelligence.
Google is shaking up Android search in response to antitrust action, Thomas Ricker reports:
Starting in early 2020, Google will present a new search provider choice screen to Android users in Europe when first setting up a new phone or tablet. The selection will then be the default search provider that powers the search box on the Android home screen as well as the Chrome browser if installed. Search providers will be required to pay Google each time a user selects them from the choice screen. Inclusion on the choice screen will be determined through a sealed-bid auction, with the top three bidders added alongside Google search.
Alex Heath had Friday’s big story: Facebook is bringing its branding to its two other big apps in an effort to burnish its own brand.
Employees for the apps were recently notified about the changes, which come as antitrust regulators are examining Facebook’s acquisitions of both apps. The app rebranding is a major departure for Facebook, which until recently had allowed the apps to operate and be branded independently. The distance has helped both apps avoid being tarnished by the privacy scandals that have hurt Facebook. The move to add Facebook’s name to the apps has been met with surprise and confusion internally, reflecting the autonomy that the units have operated under.
But Facebook CEO Mark Zuckerberg has also been frustrated that Facebook doesn’t get more credit for the growth of Instagram and WhatsApp. Associating those apps with Facebook could improve the overall companies’ brand with consumers.
This year has seen more outages of major internet platform than any I can recall. What’s going on?
Sal Rodriguez looks at the rocky early history of Facebook’s hardware efforts:
Portal sales have been so disappointing that Facebook has slashed prices multiple times. According to IDC, the company has shipped just over 54,000 Portal devices since its launch (The Information first cited this data). Michael Levin of Consumer Intelligence Research Partners described Portal’s market share and consumer awareness as “immaterial.”
Social networks grow up so fast, y’all. Here, Dami Lee highlights what has been popular there in each month of its existence.
I’m not even going to pretend to understand this one:
Superplastic, the designer toy brand launched by Kidrobot founder Paul Budnitz, has raised a $10 million Series A round of funding to turn two of its characters into animated digital media stars: Starting Tuesday, the company’s Janky and Guggimon characters will come to life on Instagram, where they will publish animated videos, interact with their fans and more.
“Our characters live their lives on social media,” Budnitz told Variety during a recent interview. “It’s like making a movie that never ends.”
Megan Graham profiles Snap’s Jeremi Gorman:
Advertisers who work with Snap are seeing the momentum. Several ad industry insiders who spoke to CNBC said Gorman’s strong reputation at Amazon preceded her as someone who’s able to drive the sales organization at a large corporation to success. During the company’s strong second-quarter earnings last week, Gorman was featured prominently.
She spoke of the company’s intention to increase the number of active advertisers using Snapchat to grow revenue, and told investors the company’s self-serve augmented reality tool has been scaling quickly. She outlined case studies with “Game of Thrones,” Subway and the oral care brand Quip. She said the company had completed its sales reorganization in the U.S. And she finished with somewhat of a zinger stat: In the second quarter of 2019, North America year-over-year average revenue per user growth rate was 42% — its highest growth rate since the second quarter of 2017.
British YouTuber Alfie Deyes marks 10 years on the platform by interviewing Wojcicki about a variety of topics in the news, including harassment and demonetization.
And finally ...
The least important question of the day also offers us a much-needed chance to laugh about something:
If you’ve logged in to Twitter today, you may have noticed that your feed is overrun with talk of hogs—30-50 feral hogs, to be specific, that run into your yard within 3-5 minutes while your small kids play.
For those confused about the abundance of hog talk, it all started when musician and friend of GQJason Isbell reacted to this past weekend’s tragic mass shootings with a tweet that read: “If you’re on here arguing the definition of ‘assault weapon’ today you are part of the problem. You know what an assault weapon is, and you know you don’t need one.” While the responses ranged from those supporting his stance to those arguing against stricter gun laws, one in particular stood out above all others. “Legit question for rural Americans,” it began. “How do I kill the 30-50 feral hogs that run into my yard within 3-5 mins while my small kids play?
Click through for many ridiculous tweets.
Talk to me
Send me tips, comments, questions, and gun control: firstname.lastname@example.org.