NewsBeat

20 NewsBeat May/June 2017 How algoriths enhance bias and work against media literacy rofessor Jennifer Pybus, a senior lecturer at the London College of Communication, studies the political economy and architecture of third party applications. In her view, one of media's main challenges can be found in the algorithmic literacy of its users: "People don't necessarily want to be challenged. If they see something that confirms their ideological position, they feel more comfortable." Producers of media services face a double challenge: first, their users should be aware of the flaws inherent in algorithms which are made by humans, may include their biases and, in the words of Vint Cerf, co-inventor of the internet protocol, "may not work exactly the way they were intended or the way P we expect them to." Secondly, how do we develop the critical thinking to understand the innards of our algorithmic digital life? Maybe Spotify highlighting music that conforms our general tastes is harmless: Discover Weekly, Related Artists and Recommended Songs all fill this user need–but in the news ecosystem, the same algorithm might enhance our confirmation bias and work against core creeds of media literacy. What do Leave.EU, #MAGA and conversion rate have common? In the "transition from mass to micro media", as Pybus describes the digitization journey of advertising, consumers benefit from micro-targeting and "appreciate being served only ads that By CLARA LLAMAS are just for them". However, the mechanics of micro-targeting are contrary to the information needs of informed citizens, who need to see all the views - not just the ones they like. "There's a missing link back to the society we live in–a tension that need exist to provide context", Pybus observes. The example helps illustrate the titanic ethical challenges in media service design. In a sense, both the "Leave.EU" and "Make America Great Again!" campaigns were service experiences. Both campaigns made innovative use of data-driven psychometric micro-targeting to find "hot- button topics that got people mad enough to get out the door and vote", in the words of professor and award-winning data-journalist Jonathan Albright. He calls this "the new data-industrial complex" and points to how a deft combination of available technologies, platforms and strategic data mining were deployed to galvanize users (in this case, passive voters). Video Killed the Radio Star — by awesome service design Of late, a trove of experience-guided digital services have been deployed inside and outside media organizations to address some of the industry's specific risks. AllSides curates center, right and left perspectives in a visually compelling way to expose bias and provide multiple angles on the same story. BuzzFeed's "Outside Your Bubble" allows users to see what people outside their social networks are saying about a news item — although not everyone's on board; the app ReadAcrossTheAisle presents a color-coded newsfeed that goes from moderate to partisan. At the bottom of the app you see a dial indicating in which part of the spectrum your daily news diet falls. Then, much like a

• Cover