The AP’s newest business reporter is an algorithm

How to upset a newsroom.
How to upset a newsroom.
Image: AP Photo/Joseph Kaczmarek
We may earn a commission from links on this page.

Journalistic earnings stories can feel robotic, even when written by a news organization as prestigious as the Associated Press. Acknowledging this fact, the AP has decided that it will just have robots produce stories on companies’ earning reports.

In an announcement on its blog, the AP announced that it will be moving toward full automation of 150- to 300-word earnings reports. The system, to be rolled out next month, will work by pumping data from Zacks Investment Research into Automated Insights, a firm that specializes in computer-generated prose. Naturally, the reports will still conform to AP style, the system of grammar and word choice that is standard in much of American journalism.

There are obvious benefits to having robots write earnings reports. For one, it facilitates a massive increase in the volume of content. The AP hopes to ramp up from providing 300 manual reports each quarter now to as many as 4,400 with the new system. An automated system also frees up reporters to work on more creative efforts, including analyzing the reports and writing stories based on them.

This is not the first time a news organization has turned to algorithms to help with repetitive writing tasks. The business information provider Thomson has been using computers to write stories since before it acquired the wire service Reuters (paywall). Many Forbes bylines belong to the Automated Insights competitor Narrative Science. The AP is already using Automated Insights to generate text on NFL players’ weekly performance on its site that ranks them. And if the Japanese robotics industry has its way, even TV news anchors could find themselves displaced by androids.

The AP says explicitly in its announcement that moving to automation won’t reduce its human reporting staff. It might, however, shift the news outlet’s hiring focus from output efficiency to analytical reasoning.

The predictability of having robots turn data into business news will certainly make life easier for one group of people: algorithmic traders developing robots that turn business news into data. Already, quantitative analysts run news stories through computer programs that parse them for information relevant to the day’s market sentiment.

Soon, instead of trying to understand humans, these programs will just have to understand other robots.