Unreported news story: 'Algorithmic takeover' of journalism |
An absolutely ridiculous news story claiming that Amazon's shopping-suggestion algorithm encourages terrorism generates an inordinate amount of internet traffic. Most of the traffic is from other news outlets republishing the claims of a single misguided source. A small percentage of the traffic is the result of stories refuting the original report's unfounded and sensationalized misstatements. A single example of the worst of journalism ends up generating revenue for several different big-name publishers.
One of those refuting the original erroneous report is Maciej Cegłowski, who writes in a September 21, 2017, post on his Idle Words blog that the claims made first in a newscast on Britain's Channel 4 were picked up by the New York Times minus one bogus claim, then repeated in whole by Vice, the Independent, and Slate. Ceglowski notes that only "skeptical BBC" took the time to check the claims in the Channel 4 report. The author of one of those subsequent stories claims it was written on a short deadline, which justifies not taking the time to verify the allegations made about the Amazon shopping-suggestion algorithm. Ceglowski believes there is an algorithm to blame, but it's not Amazon's. It's the algorithm that rewards clicks and only clicks, which has destroyed journalism. Quality doesn't matter. Accuracy doesn't matter. Clicks matter. As Ceglowski writes, "the only measure of success in publishing is whether a story goes viral on social media." The Google-Facebook duopoly requires that all stories be framed in a way that maximizes page views. Ceglowski states that the "algorithmic takeover of the public sphere is the biggest news story of the early 21st century." The problem, of course, is that there are no journalists left to cover the story. Some robots are built, others are recruited The steep decline in the ranks of journalists may soon be mirrored in other industries that are subject to a different kind of automation, in the form of robots. Brett Frischmann and Evan Selinger believe the robots have already arrived: Human workers are being transformed into robots by the gig economy. In a September 25, 2017, opinion piece on the Guardian, Frischmann and Selinger trace the roots of the "algorithmic management" that drives Uber, Deliveroo, and other digital platforms to an early 20th century management theory proposed by Frederick Taylor in "The Principles of Scientific Management." Taylorism is predicated on maximizing the value of human labor by treating workers like cogs in machines. The backlash against this management approach was swift: Critics described it as "robotism." Algorithmic management puts a modern spin on Taylorism, and while it is commonly associated with Uber and other gig-economy platforms, the philosophy has crept into the logistics and trucking industries, according to Frischmann and Selinger. Employers use sensors and other technologies to track workers' every move, and unlike in the time of Taylor when humans did the surveilling, the workers today aren't always aware of when they're being monitored. Frishmann and Selinger claim the only way to combat Taylorism is through government regulation and "strengthened workers’ rights through institutions – such as unions." Anyone within earshot of a TV or other news source knows the opposite is happening: Government regulations are being removed, and unions are being weakened. In other words, you may already be working with robots. Or maybe you're halfway to being a robot yourself. ------------------------------------------- Linkapalooza Web sites tap your electricity for payment. Online ads don't work. They don't work for site visitors, and they don't work for site owners. At least two sites have found another way to exact payment from their visitors: They tap the power of users' computers to mine bitcoins and other cryptocurrencies. The Guardian's Alex Hern writes in a September 27, 2017, article that the BitTorrent search site The Pirate Bay and the video streaming service Showtime now surreptitiously send code to the computers of their visitors that is used to mine cryptocurrencies. In addition to slowing the machines down, the code drains electricity from the computers, using up laptop batteries and sending the owners' electricity bills "soaring," according to Hern. The Pirate Bay claimed the practice was "just a test" when reporters discovered it earlier this month, while Showtime stopped sending the mining code to visitors' computers once the practice was disclosed. Hern quotes Malwarebytes analyst Jerome Segura explaining that the sites' goal of replacing the revenue generated by ads with users' bitcoin-mining assistance may be "favorable" to users, but neither company disclosed to their visitors that they were placing the cryptocurrency-mining code on their computers. Segura claims "double-dipping" by tapping power and other resources from users is a common practice. Video and gaming sites gobble up the resources of users' machines, so the overhead of processing currency-mining code is not as noticeable to users, who may be inured to the "bad user experience" that results. ------------------------------------------- Is the harm of online copyright infringement imaginary? Last week's Weekly described a new web standard that gives copyright holders more control over our access to internet content. In 2014, the European Commission paid a Dutch research firm 360,000 euros to determine the extent of the damage to copyright holders that results from online infringement. The report was delivered to the EC in May 2015, but it was never published. At least not until Julia Reda got hold of the report, as she explains in a September 20, 2017, post on her blog. Entitled "Estimating displacement rates of copyrighted content in the EU" (pdf), the study concludes that "there is no evidence to support the idea that online copyright infringement displaces sales," according to Reda. Rather than speculate about the EC's reasons for keeping the study results under wraps, Reda emphasizes the importance of providing the public with ready access to the results of public-funded research. ------------------------------------------- A free tool for reporting hate crimes is available from CuroLegal. ABA Journal's Stephanie Francis Ward reports in a September 22, 2017, article that the HateCrimeHelp.com site prompts people who believe they are the victims of a hate crime to answer a series of questions about the incident -- where it occurred, under what circumstances -- and then serves up information about how to report the crime to local authorities. Past Weeklies have linked to ProPublica's hate crime scorecard, which the nonprofit news organization created following the 2016 Presidential election. The need for diligent reporting on hate crimes is shown by the results of a study (pdf) written by Prof. Brian Levin of the Center for the Study of Hate and Extremism at California State University, San Bernardino. Christopher Mathias of the Huffington Post reports in a September 17, 2017, article that after collecting data from 31 large cities and counties in the U.S., the researchers found the number of hate crimes increased 5 percent overall in 2016 from 2015. Hate crimes were reported 50 percent more often in Philadelphia, 20 percent in Chicago, 24 percent in New York City, 15 percent in Los Angeles, and a whopping 62 percent in Washington, D.C. In nearly all areas surveyed, there was a spike in hate-crime reports in the weeks leading up to the November 2016 Presidential election. ------------------------------------------- Apple's bulk data collection respects your privacy... or so says Apple. Some web sites don't play by the rules. They may be power-hungry (as noted in the item above), or they may attempt to grab so much of your system's memory that your browser crashes. The version of the Safari browser in Apple's iOS 11 and macOS High Sierra attempts to identify misbehaving sites by using a technique called differential privacy to scarf up mass quantities of data about how people browse in a way that doesn't disclose personal information, according to Apple. TechCrunch's Brian Heater reports in a September 25, 2017, article that Apple previously used differential privacy in the predictive text of keyboards, search predictions, and selection of emojis. Safari users have to opt in to allow their browser data to be sent to Apple, similar to the opt-in option for reporting crashes. As Heater points out, participating in the differential privacy program requires buying into the notion that collecting more of your personal data will somehow improve your privacy. (Note that researchers at Cornell University concluded that Apple's differential privacy isn't as private as the company claims. In a paper first submitted on September 7, 2017, the researchers call for more transparency from Apple about its implementation of differential privacy, as well as more choice for users to restrict the collection of their personal data, and greater care in setting the system's privacy "default.") ------------------------------------------- Improve your health by hitting the hay. Imagine leaving your next doctor visit with a prescription for eight hours of sleep, seven nights a week. Matthew Walker, who is the director of the Center for Human Sleep Science at the University of California, Berkeley, claims there is a "catastrophic sleep-loss epidemic" that poses a serious public health risk. Walker's new book, "Why We Sleep," ties sleep loss to such maladies as Alzheimer’s disease, cancer, diabetes, obesity, and mental illness. Walker is profiled by the Guardian's Rachel Cooke in a September 24, 2017, article. The researcher points out that in 1942, fewer than 8 percent of the U.S. population tried to get by on six hours of sleep or less. In 2017, about half of us manage six or fewer hours of sleep each night. Since seeing in his research results the detrimental effects of sleep deprivation on his subjects, Walker now gives himself a "non-negotiable eight-hour sleep opportunity every night." ------------------------------------------- I'd like to take this non-negotiable opportunity to bid you a good night (or day, as the case may be). What do you say we make a pact: For the next week, we'll spend one fewer hour each day in front of a screen, and one more hour each night in dreamland. Live long and slumber. (Or is that slumber and live long?) |
