Artificial Things

Inside a Twitter bot farm; is AI the new oil?; and more


  • This is civic tech: A combination of high-tech satellites and lower-tech sousveillance is beginning to enable the policing of illegal timber sales, Richard Conniff reports for The New York Times. While the problem of illegal logging is global, in Romania the Environment Ministry has released an app called Forest Inspector that encourages ordinary citizens to report logging trucks. Within days of its release, 30,000 people filed reports and “overnight, the number of truck applying for permits and registering their loads increased by 50 percent,” he reports.(h/t Jonathan Fox)

  • Zach Graves and Kevin Kosar of R Street, a center-right libertarian think tank, have written a paper arguing that Congress should bring back the nonpartisan Office of technology Assessment (axed by Newt Gingrich) because the legislature needs help keeping pace with technological innovation.

  • Tech and politics must-read: The New York Times’ big investigation of the black market for Twitter followers, written by Nicholas Confessore, Gabriel J.X. Dance, Richard Harris, and Mark Hansen. The paper shows how it is possible to “fingerprint” a person’s Twitter follower list in order to spot when huge blocs of new followers suddenly add to their total. (Here’s an example of how that can be visualized.)

  • By some accounts, as many as 48 million of Twitter’s reported active users, nearly 15 percent, are automated accounts designed to simulate real people, the Times reports. One free tool that you can use to get a sense of your own fake Twitter followers is TwitterAudit.com.

  • It’s worth remembering that people with large followings on Twitter have often had artificially inflated numbers. In Twitter’s early days that came from the company itself, which gave favored friends huge followings by putting them on a “suggested users” list. This included lots of people in the tech industry who were genuine friends of the founding group as well as “influencers” whose support they wanted to cultivate. Developer and blogger Dave Winer called them out for it, asking “Why is Tim O’Reilly on the list, but not Jay Rosen? Why TechCrunch and not GigaOm, PaidContent, SiliconAlley, ReadWriteWeb, Mashable, VentureBeat, etc etc.”

  • Data engineer Hilary Mason explains that when she started seeing bots pop up copying her persona, she wrote her own bot army to fill up the namespace and block anyone else from impersonating her.

  • New York state attorney general Eric Schneiderman is opening an investigation into Devumi, the firm that sold millions of fake followers on Twitter, Nicholas Confessore reports. “Impersonation and deception are illegal under New York law,” Schneiderman says. We’re going to need a lot more enforcement of such laws.

  • Related: Russian-linked Twitter bots retweeted Donald Trump nearly half a million times in the weeks before the election in 2016, Twitter has told the Senate Judiciary Committee in a written statement—amounting to 4.25  percent of all retweets of his account in that period. Just .55 percent of Hillary Clinton’s tweets were retweeted by these accounts during the same period.

  • The Trump administration is considering nationalizing the country’s 5G wireless network to harden it against Chinese spying, Jonathan Swan reports for Axios.

  • What sharing economy? In her latest Big City column, The New York Times’ Gina Bellafante takes a hard whack at WeWork for encouraging a bro-ey atmosphere in what is supposed to be a workspace, and tops that off with a jab at AirBnb for making a few rich hosts a lot richer.

  • A group of current Google employees say right-wing coworkers are inciting outsiders to harass Googlers who are minority advocates, including queer and transgender employees, Natasha Tiku reports for Wired. Screenshots from internal company discussion forums are showing up on sites like Breitbart, leading to waves of harassment generated from forums like 4chan and Kiwi Farms.

  • Technology critic Evgeny Morozov asks a good question in his latest Guardian column: How long will tech firms need us (their users) to train their AI systems? As he argues, “data is not the new oil, …but AI probably is.”

  • Privacy, shmivacy: To celebrate International Data Privacy Day, the good folks at Privacy International are spending all week highlighting how countries are addressing data governance.

  • Related: Fitness app Strava has released their global heat map with 3 billion GPS points from their users, and a number of security and military geeks have pointed out that this is an excellent way to spot military bases, including supposedly secret ones. More here from Liz Sly of the Washington Post.

  • As academic Henry Farrell writes in a trenchant essay in the Boston Review that I think I will just point to again and again as this crazy stuff keeps getting more normal: We live in a Philip K. Dick future, “a world in which technology is developing in ways that make it increasingly hard to distinguish human beings from artificial things.”

  • Apply: Stanford’s Philanthropy and Civic Society program is looking to hire a program manager to support its Project on Democracy & Internet, which is being funded by the Knight Foundation and the Hewlett Cyber Initiative.