Many Americans — and possibly even non-Americans — have probably had enough of this year’s presidential election coverage. Interestingly, some of the election news written over the past few months wasn’t created by trained journalists — at least not trained in the traditional sense. “Bots” or “robo-journalists” have actually provided content for some of the top publications in the United States.
The bots’ employers include the New York Times, Washington Post, CNN, NBC, Yahoo News, and the non-profit ProPublica, according to the Agence France-Presse. Each of these publications use algorithms to sift through data and turn it into text, or chatterbots to deliver news coverage to readers via messaging apps.
“This is a natural follow-on to what we have been doing in conversational journalism,” New York Times product director Andrew Phelps told AFP.
Earlier this year, the New York Times began using a bot for reporter Nick Confessore that reaches out to readers via Facebook Messenger. When the bot first makes contact, it does so through a short, punchy phrase like “Hey it’s Nick. The race took a swerve this weekend.” The reader is then encouraged to engage further with the story. Rather than attempting to pull in immediate revenue, these bots are expected to build connections and relationships with the audience.
Meanwhile, nonprofits like ProPublica and the PollyVote Project are using algorithms to turn data into text, and are delivering updates to coverage, forecasts, trends, and campaign finance reports every 15 minutes.
The team over at the Washington Post has developed Heliograf, a system that produces stories written by a combination of humans and computers in an effort to automate while maintaining quality. Tonight, the publication will use an artificial intelligence to updates dispatches, according to AFP.
But bots aren’t exclusive to the United States. A series of European studies suggest that readers there are often oblivious to whether text was written by a human or computer.
“When you ask people how readable a story is, they rate the human article better, but when you ask them how credible it is, the computer is better,” said Andreas Graefe, a fellow at Columbia University’s Tow Center for Digital Journalism who is leader of a related project that is funded by the Tow Center and Germany’s LMU Munich University. “We don’t really know why.”
Robo-journalists aren’t expected to replace reporters soon but they’re already helping organizations like the Associated Press and Bloomberg craft data-driven reports on minor-league sports and finance, so it’s only a matter a time before these systems become more prolific in the newsroom.