Christer Clerwall (Karlstad University, Sweden) presented a total of 45 test subjects with texts that had been generated partly by computers and partly by journalists. Respondents rated the machine-generated texts as slightly "more informative," "more accurate," "more trustworthy," and "more objective," but also "more boring," while the journalists scored high on "coherence," "clarity," and "reading pleasure" - but overall these differences were "not significant."

Hille van der Kaa and Emiel Krahmer (both Tilburg University in the Netherlands) specifically investigated the credibility of "robot" texts and surveyed 232 Dutch people, including 64 journalists. Interestingly, the ordinary citizens surveyed gave computers and journalists equally good marks in terms of credibility and expertise. Only the journalists attested more trustworthiness to their own profession than to their robot colleagues. But this is probably only the tip of the iceberg. In times when there are already four to five PR people for every journalist in the U.S., selecting from the media releases supplied is even more important than writing articles. Algorithms are also increasingly taking over this selection in search engines and social networks, and media experts such as Caitlin Dewey (Washington Post) and Walter Quattrociocchi (Institute for Advanced Studies, Lucca) are worried that our view of the world could become very one-sided in this way - simply because the algorithms only provide us with what we want to know, and not what we should know.

Nikki Usher (George Washington University), however, gives the all-clear. In her view, the problem of "selective perception," i.e., the fact that we prefer to consume media offerings that confirm our prejudices, is and remains our problem as media users and not that of the search engines and social networks that supply us with information. Their algorithms are much more complex than the Cassandra callers generally make them out to be. For example, the Google News algorithm favors certain news items, such as those provided by large editorial departments like The New York Times, or offers that contain more context, for example in the form of original quotes, or even particularly current news. In principle, the people responsible for Google News would strive to offer the "best possible news selection" in a similar way to professional journalists - only with the help of algorithms that respond to user interests.

Critics often overlook what "algorithms that filter news have in common with conventional news selection processes": "For a long time, newsrooms were also something like 'black boxes' - staffed by journalists who decided on the top stories of the day behind impenetrable walls." Nonetheless, of course, the door remains open for manipulation and remote control when giant corporations like Google or Facebook decide centrally on the use of algorithms - and can thus program worldwide what Chen Cheng in Singapore, John Sixpack in Tennessee and Mrs. Hinz and Mr. Kunz in Appenzeller Land get to see, hear and read. Nor do we necessarily share the optimism of some researchers that the use of robots in newsrooms could free up resources for more important tasks such as research. Clerwall sees it more realistically as "a way for media companies to save on staff."

For quite some time, there has been a twilight of the gods in editorial offices: For decades, journalists were used to reporting only on the unemployment of others without being affected by it themselves. Since the turn of the century, at the latest, this has changed thoroughly. Computer programs and algorithms are likely to dramatically accelerate job losses in editorial departments once again - and then presumably not even stop at the editorial suppliers, the PR industry.

Stephan Russ-Mohl directs the European Journalism Observatory at the University of Lugano.

Literature tips:

Christer Clerwall (2014): Enter the Robot Journalist, in Journalism Practice 8:5, 519-531. http://www.tandfonline.com/doi/abs/10.1080/17512786.2014.883116?journalCode=rjop20#.VZJDR6QwKH4

Hille van der Kaa/Emiel Krahmer (2014): Journalist versus news consumer: The perceived credibility of machine written news, Tilburg University. https://pure.uvt.nl/portal/files/4314960/cj2014_session4_paper2.pdf

Caitlin Dewey (2015): http://www.washingtonpost.com/news/the-intersect/wp/2015/06/03/if-you-use-facebook-to-get-your-news-please-for-the-love-of-democracy-read-this-first/

Nikki Usher (2015): Who's afraid of a big bad algorithm? In: Columbia Journalism Review,.June 11,http://www.cjr.org/analysis/whos_afraid_of_a_big_bad_algorithm.php

Christer Clerwall (Karlstad University, Sweden) presented a total of 45 test subjects with texts that had been generated partly by computers and partly by journalists. Respondents rated the machine-generated texts as slightly "more informative," "more accurate," "more trustworthy," and "more objective," but also "more boring," while the journalists scored high on "coherence," "clarity," and "reading pleasure" - but overall these differences were "not significant."

Hille van der Kaa and Emiel Krahmer (both Tilburg University in the Netherlands) specifically investigated the credibility of "robot" texts and surveyed 232 Dutch people, including 64 journalists. Interestingly, the ordinary citizens surveyed gave computers and journalists equally good marks in terms of credibility and expertise. Only the journalists attested more trustworthiness to their own profession than to their robot colleagues. But this is probably only the tip of the iceberg. In times when there are already four to five PR people for every journalist in the U.S., selecting from the media releases supplied is even more important than writing articles. Algorithms are also increasingly taking over this selection in search engines and social networks, and media experts such as Caitlin Dewey (Washington Post) and Walter Quattrociocchi (Institute for Advanced Studies, Lucca) are worried that our view of the world could become very one-sided in this way - simply because the algorithms only provide us with what we want to know, and not what we should know.

Nikki Usher (George Washington University), however, gives the all-clear. In her view, the problem of "selective perception," i.e., the fact that we prefer to consume media offerings that confirm our prejudices, is and remains our problem as media users and not that of the search engines and social networks that supply us with information. Their algorithms are much more complex than the Cassandra callers generally make them out to be. For example, the Google News algorithm favors certain news items, such as those provided by large editorial departments like The New York Times, or offers that contain more context, for example in the form of original quotes, or even particularly current news. In principle, the people responsible for Google News would strive to offer the "best possible news selection" in a similar way to professional journalists - only with the help of algorithms that respond to user interests.

Critics often overlook what "algorithms that filter news have in common with conventional news selection processes": "For a long time, newsrooms were also something like 'black boxes' - staffed by journalists who decided on the top stories of the day behind impenetrable walls." Nonetheless, of course, the door remains open for manipulation and remote control when giant corporations like Google or Facebook decide centrally on the use of algorithms - and can thus program worldwide what Chen Cheng in Singapore, John Sixpack in Tennessee and Mrs. Hinz and Mr. Kunz in Appenzeller Land get to see, hear and read. Nor do we necessarily share the optimism of some researchers that the use of robots in newsrooms could free up resources for more important tasks such as research. Clerwall sees it more realistically as "a way for media companies to save on staff."

For quite some time, there has been a twilight of the gods in editorial offices: For decades, journalists were used to reporting only on the unemployment of others without being affected by it themselves. Since the turn of the century, at the latest, this has changed thoroughly. Computer programs and algorithms are likely to dramatically accelerate job losses in editorial departments once again - and then presumably not even stop at the editorial suppliers, the PR industry.

Stephan Russ-Mohl heads the European Journalism Observatory at the University of Lugano.

Literature tips:

Christer Clerwall (2014): Enter the Robot Journalist, in Journalism Practice 8:5, 519-531. http://www.tandfonline.com/doi/abs/10.1080/17512786.2014.883116?journalCode=rjop20#.VZJDR6QwKH4

Hille van der Kaa/Emiel Krahmer (2014): Journalist versus news consumer: The perceived credibility of machine written news, Tilburg University. https://pure.uvt.nl/portal/files/4314960/cj2014_session4_paper2.pdf

Caitlin Dewey (2015): http://www.washingtonpost.com/news/the-intersect/wp/2015/06/03/if-you-use-facebook-to-get-your-news-please-for-the-love-of-democracy-read-this-first/

Nikki Usher (2015): Who's afraid of a big bad algorithm? In: Columbia Journalism Review,.June 11,http://www.cjr.org/analysis/whos_afraid_of_a_big_bad_algorithm.php