After an earthquake hit Los Angeles Monday morning, it took only three minutes for the LA Times to write and publish an article about it online. An algorithm from programmer-journalist Ken Schwenke is designed to automatically generate a quick article when an earthquake occurs, and it could point to the future of digital journalism.
The bot, known as Quakebot, draws on sources like the U.S. Geological Survey for data and inserts the numbers into a written template, Slate Magazine explained. All Schwenke had to do was give a quick look-over and click “publish.”
The LA Times also has bots that can generate general crime stories, and other news organizations have used bots to generate short reports on sports games.
Clearly, there may be pitfalls with relying on bots to publish information. Earlier this month, an investigator found that more than 120 research papers published by the Institute of Electrical and Electronic Engineers and Springer were fake and generated by an algorithm developed in 2005 by two MIT students as a prank.
This is why these “robo-journalists” aren’t designed to replace human journalists, Schwenke says. They are intended to be tools that can gather and assemble relevant information to get the basic facts out to the public faster.
“It saves people a lot of time, and for certain types of stories, it gets the information out there in usually about as good a way as anybody else would,” he told Slate magazine, adding that he doesn’t believe bots will eliminate jobs.
Quakebot cannot interview experts or provide relative context. After the initial report, various human writers and editors at LA Times took Quakebot’s initial story and turned into it an in-depth piece for the front page.
A robot journalist won’t write a Pulitzer-caliber feature piece anytime soon, but it will help human journalists have more time to do so.