Are Algorithms Replacing Reporters?

Artificial intelligence developers are supplying major news outlets with story-producing software

While Wall Street companies have been relying on algorithm-driven news for several years, utilizing its real-time data reports in making investment decisions, top news agencies have recently begun using algorithmic software, or automation technology, to generate and publish narrative stories electronically. In the past few years, Automated Insights and Narrative Science have allowed news companies to rapidly produce prose-style articles.

Unlike more traditional automation technology, which sifts through massive amounts of data, detects likely areas of interest, and offers up spreadsheets in summary, programs like Narrative Science’s Quill aim to thoroughly digest information and respond to directives with thoughtful analyses written in (for now) plain English, a process dubbed “narrative analytics.” As Narrative Science’s chief scientist and co-founder Kris Hammond tells Business Insider, “[w]e saw the beginnings of dissatisfaction with big data and we saw ourselves as the solution. You don’t want spreadsheets, you want to be told.” Quill’s ability to both gather and narrate data is put to perfect use with GameChanger, a mobile app and website for managing fantasy sports team rosters, automatically assembling player stats, and providing live updates as well as narrative recaps of games.

algorithm 4.jpg

Quill software also adapts its research algorithms to the individual needs of its target members and communities, learning “what really matters to your business” and “automatically applying language to the most relevant insights.” This push for personalized content is shared by the developers of Automated Insights’ Wordsmith, who propose an entirely new perspective on, and set of expectations from, the future shape of news. “We flipped the standard content creation model on its head,” Automated Insights CEO Robbie Allen explained to Poynter in March. “The standard way of creating content is, ‘I hope a million people read this.’ Our model is the inverse of that. We want to create a million pieces of content with one individual reading each copy.”

For groups seeking to reach mass audiences, too, the software is popular. Thomson Reuters has been using similar technology since 2006, while the LA Times used their QuakeBot to report an earthquake within a few minutes of its happening in March. Taking a major step in joining this trend, the Associated Press, which was previously employing automation technology to report sports statistics, has announced that it will begin using Automated Insight’s technology to produce the majority of its corporate earnings stories. In the June 30th, 2014 press release on the company’s website, AP’s managing editor Lou Ferrara explains the multinational news agency’s decision to rely on automation to produce earnings reports, emphasizing efficiency:

We are going to use our brains and time in more enterprising ways during earnings season. Rather than spending a great deal of time focusing on the release of earnings and hammering out a quick story recapping each one, we are going to automate that process for all U.S. companies in the 4,400.

Like other news outlets, the Associated Press currently remains committed to the idea of retaining its human staff for higher-level analysis, their pursuits supplemented by automated content. Experiments in other directions do exist, however, as in the case of a Japanese news-reading robot.

Find samples of automated narratives (to analyze for yourself) via Slate Magazine and Business Insider.

[h/t] Poynter, Quartz

Images: NPR, KCBS

Quantcast