Automated artificial intelligence journalism, now commonly called ‘Robo-Journalism’ is likely to expand throughout the industry in spite of its current limitations according to a research project at German and Swiss universities.
Journalists and editors believe ‘robo-journalists’ do not have a good nose for news and produce one-dimensional stories, according to new research published by researchers at Munich and Zurich universities.
However, despite the limitations, the report reveals plans for the technology to be rolled out more widely with the potential to replace “hundreds” of journalists at Thomson Reuters alone.
The researchers, Neil Thurman (Ludwig-Maximilians-Universitaet (LMU) in Munich), Konstantin D?rr (University of Zurich) and Jessica Kunert (LMU Munich), interviewed journalists, editors, and executives from CNN, BBC, Thomson Reuters, Trinity Mirror, and News UK in an exploratory study. The journalists were given hands-on experience with robo-writing software during a workshop.
Robo -- or automated -- journalism, is software that converts structured data into stories with limited to no human intervention beyond the initial programming. It is used by news organizations including Associated Press, the Los Angeles Times, and Forbes.
The report was published in the international peer-reviewed journal, Digital Journalism on March 1.
The journalists and editors in Thurman’s study believe robo-journalism’s reliance on data streams and the need to program news angles in advance means the stories produced lack the context, complexity, and creativity of much traditional reporting.
Journalists also thought the need to template robo-written stories in advance is a drawback.
One journalists from the BBC, said “you cannot get a reaction to those numbers, you cannot explain or interrogate them because the story template was written before the numbers came out” and concluded, after using robo-writing technology first hand, it was not worth the BBC researching the technology further.
Despite these shortcomings, journalists do believe robo-journalism does have the potential to reduce costs and increase the speed and specificity of some reporting.
Journalists at CNN and Reuters thought it could “reduce costs” by replacing “expensive staff” who are doing “fairly simplistic and time-consuming work”.
A Reuters journalist believed automation could improve speed and accuracy, and said “we are looking at it in all parts of the company”.
Another Reuters journalist said automation will be used for stories that do not “have the resources to cover manually” or for topics currently below the threshold of reportability.
Add new comment
Read our comment policy before posting your viewpoints