So­cial me­dia’s mis­in­for­ma­tion bat­tle: No win­ners, so far

The Progress-Index Weekend - - OBITUARIES - By Bar­bara Or­tu­tay

NEW YORK — Face­book and other so­cial plat­forms have been fight­ing on­line mis­in­for­ma­tion and hate speech for two years. With the U.S. midterm elec­tions just a few days away, there are signs that they’re mak­ing some head­way, al­though they’re still a very long way from win­ning the war.

That’s be­cause the ef­fort risks run­ning into po­lit­i­cal head­winds that Face­book, Twit­ter and Google find bad for busi­ness. Some even ar­gue that the so­cial net­works are easy to flood with dis­in­for­ma­tion by de­sign — an un­in­tended con­se­quence of their ea­ger­ness to cater to ad­ver­tis­ers by cat­e­go­riz­ing the in­ter­ests of their users.

Caught em­bar­rass­ingly off-guard af­ter they were played by Rus­sian agents med­dling with the 2016 U.S. elec­tions, the tech­nol­ogy giants have thrown mil­lions of dol­lars, tens of thou­sands of peo­ple and what they say are their best tech­ni­cal ef­forts into fight­ing fake news, pro­pa­ganda and hate that has pro­lif­er­ated on their dig­i­tal plat­forms.

Face­book, in par­tic­u­lar, has pulled a ma­jor re­ver­sal since late 2016, when CEO Mark Zucker­berg in­fa­mously dis­missed the idea that fake news on his ser­vice could have swayed the elec­tion as “pretty crazy.” In July, for in­stance, the com­pany an­nounced that heavy spend­ing on se­cu­rity and con­tent mod­er­a­tion, cou­pled with other busi­ness shifts, would hold down growth and prof­itabil­ity. In­vestors im­me­di­ately pan­icked and knocked $119 bil­lion off the com­pany’s mar­ket value. The so­cial net­work has started to see some pay­off for its ef­forts. A re­search col­lab­o­ra­tion be­tween New York Univer­sity and Stan­ford re­cently found that user “in­ter­ac­tions” with fake news sto­ries on Face­book, which rose sub­stan­tially in 2016 dur­ing the pres­i­den­tial cam­paign, fell sig­nif­i­cantly be­tween the end of 2016 and July 2018. On Twit­ter, how­ever, the shar­ing of such sto­ries con­tin­ued to rise over the past two years.

A sim­i­lar mea­sure from the Univer­sity of Michi­gan’s Cen­ter for So­cial Me­dia Re­spon­si­bil­ity dubbed the “Iffy Quo­tient “— which gauges the preva­lence of “iffy” ma­te­rial on so­cial net­works — also shows that Face­book’s “iffi­ness” has fallen from a high of 8.1 per­cent 1n March 2017 to 3.2 per­cent on Mon­day. Twit­ter iffi­ness has also fallen slightly, from 5.6% in Novem­ber 2016, to 4.2 per­cent on Mon­day.

Even at these lev­els, fake news re­mains huge and may be spread­ing to new au­di­ences. A team led by Philip Howard, the lead re­searcher on Ox­ford’s Com­pu­ta­tional Pro­pa­ganda ef­fort, looked at sto­ries shared on Twit­ter dur­ing the last 10 days of Septem­ber 2018 and found that what it called “junk news” ac­counted for a full quar­ter of all links shared dur­ing that time — greater than the num­ber of pro­fes­sional news sto­ries shared dur­ing that time.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.