Threats

If you read one thing today, make it this: “There is no conceivable world in which enough bomb-making equipment is being sold on Amazon to train an algorithm to make this recommendation.” So writes Maciej Ceglowski in a typically trenchant commentary on the “moral panic” taking place over a British TV station’s report that Amazon’s recommendation engine was helping people make bombs. The story, by Channel 4, has spread widely across the web, without anyone asking, as Ceglowski does, if there were more innocent reasons to buy chemicals online (say, to make fireworks) or “why on earth would an aspiring bomber use an online shopping cart tied to their real identity?”

Ceglowski adds, “The real story in this mess is not the threat that algorithms pose to Amazon shoppers, but the threat that algorithms pose to journalism. By forcing reporters to optimize every story for clicks, not giving them time to check or contextualize their reporting, and requiring them to race to publish follow-on articles on every topic, the clickbait economics of online media encourage carelessness and drama….The real story of machine learning is not how it promotes home bomb-making, but that it’s being deployed at scale with minimal ethical oversight, in the service of a business model that relies entirely on psychological manipulation and mass surveillance. The capacity to manipulate people at scale is being sold to the highest bidder, and has infected every aspect of civic life, including democratic elections and journalism.” B’doing!

Life in Facebookistan: Technosociologist Zeynep Tufecki reminds us in The New York Times that the latest Facebook scandal is just a reminder of the company’s structural strengths and flaws. “[Its] business model — ad-targeting through deep surveillance, emaciated work force, automation and the use of algorithms to find and highlight content that entice people to stay on the site or click on ads or share pay-for-play messages — works. The trouble is Facebook’s business model is structurally identical whether advertisers are selling shoes, politics or fake diet pills, and whether they’re going after new moms, dog lovers or neo-Nazis.”

It looks like it took a personal intervention from President Obama, nine days after the election, to get Facebook CEO Mark Zuckerberg to start taking the threat of disinformation being spread by his platform more seriously, as Adam Entous, Elizabeth Dwoskin and Craig Timberg report in The Washington Post.

BuzzFeed’s Charlie Warzel writes that Facebookistan is more like “the United Nations — a group of individuals endowed with the almost impossible responsibility of policing a network of interconnected autonomous powers.”

The Cyber: If President Trump ends the nuclear deal with Iran, the result could be big new round of Iranian cyber-attacks on the United States, Eric Geller writes for Politico.

Men Going Their Own Way is an all-male separatist movement that hates women, and as Nellie Bowles reports for The New York Times, it appears to be gaining ground as the backlash against gender equity grows at many Bay Area tech companies. Some of the men quoted in Bowles’ story aren’t very smart. Here’s one James Altizer, an engineer at chip maker Nvicia, who hosts private Facebook pages devoted to “mens rights.” He tells Bowles “I’m witting in a soundproof booth right now because I’m afraid someone will hear me.” Uh, you are being quoted by name in The New York Times, bozo!

What happened? Democratic pollster Stanley Greenberg writes in The American Prospect that Hillary Clinton lost the 2016 election because “The campaign relied far too heavily on something that campaign technicians call ‘data analytics.’ This refers to the use of models built from a database of the country’s 200 million–voters, including turnout history and demographic and consumer information, updated daily by an automated poll asking for vote preference to project the election result. But when campaign developments overtake the model’s assumptions, you get surprised by the voters—and this happened repeatedly. Astonishingly, the 2016 Clinton campaign conducted no state polls in the final three weeks of the general election and relied primarily on data analytics to project turnout and the state vote. They paid little attention to qualitative focus groups or feedback from the field, and their brief daily analytics poll didn’t measure which candidate was defining the election or getting people engaged.”

He also writes that “The models from the data analytics team led by Elan Kriegel got the Iowa and Michigan primaries badly wrong, with huge consequences for the race. Why were they not then fired?” It’s been almost a year since the election and Kriegel has still not be seen nor heard from.

The Department of Homeland Security has informed election officials in 21 states that hackers possibly connected to Russia tried to break into their systems during the 2016 election, the Associated Press’ Geoff Mulvihill and Jake Pearson report. The states targeted were Alabama, Alaska, Arizona, California, Colorado, Connecticut, Delaware, Florida, Illinois, Iowa, Maryland, Minnesota, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Texas, Virginia, Washington, and Wisconsin.

This is civic tech: Ory Okolloh of the Omidyar Network explains their new $180,000 grant to South Africa’s Grassroot, a nonprofit organization that develops simple and powerful collective organizing tools that don’t require smartphones or data plans.

Carrie Bishop, San Francisco’s chief digital services officer, reports on how her team has been doing.

Sleeping Giants, the group of anonymous online activists who are chipping away at Breitbart’s advertising base, gets profiled by Paul Farhi in The Washington Post. The group says it has convinced nearly 2,900 companies to pull their advertising from the alt-right site.