The most problematic aspect of Facebook’s power is Mark’s unilateral control over speech. There is no precedent for his ability to monitor, organize and even censor the conversations of two billion people.

Facebook engineers write algorithms that select which users’ comments or experiences end up displayed in the News Feeds of friends and family. These rules are proprietary and so complex that many Facebook employees themselves don’t understand them.

In 2014, the rules favored curiosity-inducing “clickbait” headlines. In 2016, they enabled the spread of fringe political views and fake news, which made it easier for Russian actors to manipulate the American electorate. In January 2018, Mark announced that the algorithms would favor non-news content shared by friends and news from “trustworthy” sources, which his engineers interpreted — to the confusion of many — as a boost for anything in the category of “politics, crime, tragedy.”

Facebook has responded to many of the criticisms of how it manages speech by hiring thousands of contractors to enforce the rules that Mark and senior executives develop. After a few weeks of training, these contractors decide which videos count as hate speech or free speech, which images are erotic and which are simply artistic, and which live streams are too violent to be broadcast. (The Verge reported that some of these moderators, working through a vendor in Arizona, were paid $28,800 a year, got limited breaks and faced significant mental health risks.)