Nieman Foundation at Harvard
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
ABOUT                    SUBSCRIBE
March 14, 2019, 11:42 a.m.
Reporting & Production
LINK:  ➚   |   Posted by: Laura Hazard Owen   |   March 14, 2019

Reports about the use of automated news in journalism usually tend to include some kind of “relax-your-job-isn’t-being-taken-by-a-machine” disclaimer, and a report from WAN-IFRA out this week is no exception.

The report focuses on “the automated generation of news texts based on structured data.” The point of this type of automated content is to save journalists from repetitive tasks, “while increasing output volume.” Machines are definitely not replacing journalists, the report’s authors say, so you can stop worrying about that. The biggest question is more mundane: “Whether the [news automation] system should be bought from a service provider or created and modified in-house.”

The report looks at five publishers from around the world that are using automated reporting. The Washington Post, for instance, has Heliograf. Originally created for the 2016 Rio Olympics, Heliograf now encompasses other areas with a lot of data, “such as election results, crime, real estate, or earnings announcements.” The Post produced 850 articles with Heliograf between 2016 and 2017.

In the U.K., RADAR (Reporters and Data and Robots) is producing data-driven local news stories (with funding from Google’s Digital News Initiative).

Journalists at Urbs Media write text templates and the system chooses the data for the location of the news outlet, which means that each template can be used for hundreds of different stories.

Thanks to the use of a market-available product, RADAR started production of stories for a pilot group of around 40 newspaper titles that turned into a full, UK-wide trial. In early 2019, RADAR became a subscription-based local news agency mainly serving large- and medium-size local news publishers.

Data from the U.K.’s National Health Service, for instance, can be used to customize a couple hundred localized versions of a single story.

And Sweden’s MittMedia is the part owner of United Robots, which started out producing news reports about team sports but has moved to real estate and bankruptcy.

For sports data, MittMedia and United Robots rely on Everysport Media Group, a closely connected company that provides the input for news. Everysport employees actually call team leaders and referees by phone after each game or collect the match facts online if possible in order to get scores from all matches played in Sweden on all levels. This is a manual, laborious and tedious way of creating your own data, but since most lower division sports organizations lack a common reporting system, this is the only feasible solution.

Recently, the companies have introduced a chat bot that collects usable quotes from team leaders. MittMedia and United Robots are constantly scanning open data in search of new opportunities. In early 2018 United Robots and MittMedia launched a system that writes news articles about bankrupted companies. In 2017, they also started to produce automated news about property deals, a service that is popular and one of the main news items converting free readers to subscribers.

Sören Karlsson, CEO of United Robots, suggested that news organizations can use automated sytems to alert them when something unusual or newsworthy is going on, “for example, when the most expensive house on the market is sold.”

“Data analysis is an important part of automated processes,” he said, “and algorithms are much better equipped to find hidden relationships, outliers, and so on, than people.”

The full report is free for WAN-IFRA members, or can be purchased from WAN-IFRA’s site here.

Illustration of a futuristic city, where presumably at least the cops beat has been taken over by robots, by James Boast used under a Creative Commons license.

Show tags
Join the 60,000 who get the freshest future-of-journalism news in our daily email.
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
Within days of visiting the pages — and without commenting on, liking, or following any of the material — Facebook’s algorithm recommended reams of other AI-generated content.
What journalists and independent creators can learn from each other
“The question is not about the topics but how you approach the topics.”
Deepfake detection improves when using algorithms that are more aware of demographic diversity
“Our research addresses deepfake detection algorithms’ fairness, rather than just attempting to balance the data. It offers a new approach to algorithm design that considers demographic fairness as a core aspect.”