by Yvonne Brown
It’s very likely many of the news stories you skimmed on your phone today while waiting for the kettle to boil or riding the bus, were written by a robot.
The recap of yesterday’s game, the financial earnings statement, the report on a natural disaster, or a summary of last night’s council meeting — may all have been put together by artificial intelligence instead of a human journalist. Newsrooms around the world facing mounting pressure to do more with less are increasingly turning to newsbots to help generate the data-rich, constant news coverage consumers are demanding.
These bots are online software applications that automate simple and repetitive tasks. The Washington Post uses a newsbot called Heliograf. Editors there create narrative templates for stories, including key phrases that account for a variety of potential outcomes (from an election for example, “Republicans retained control of the House” to “Democrats regained control of the House”). They then hook Heliograf up to any source of structured data. The software identifies the relevant information, matches it with the corresponding phrases in the template, merges them, and then publishes different versions of the story across a variety of platforms. The system can also alert reporters to any anomalies it finds in the data so they can investigate.
USA Today’s Wibbitz creates short videos by condensing news articles into a script, stringing together a selection of images or video footage, and sometimes adding narration with a synthesized voice. Reuters’ algorithmic prediction tool News Tracer helps journalists gauge the integrity of a tweet. BuzzFeed’s BuzzBot collects information from on-the-ground sources at news events. One of the first news algorithm’s was the LA Times’ Quakebot, which began writing stories on seismic activity in 2014. Interestingly, the story has a byline, but it’s not for a writer but rather the program’s designer.
In June 2018, 69 per cent of media leaders surveyed by the Reuters Institute for the Study of Journalism use some type of artificial intelligence (AI) in their newsrooms.
The pluses of bots for resource challenged newsrooms are many. AI frees time-strapped reporters from routine tasks allowing them to concentrate on more in-depth work. When there is a lot of complex data to sift, bots maybe able to get through more volume more accurately. Bots can provide the kind of hyperlocal journalism that audiences are interested in but few publishers can afford. Finally, there are all the advantages of AI personalising content, improving customer relations, and supporting journalists in discovering new stories and managing their material.
But bots come with serious downsides and not just for the journalists whose jobs they may potentially usurp. Audiences spoon fed their news by bots in snippets of facts and figures aren’t getting the full picture. The world is complicated, complex stories require in-depth analysis, careful writing and sensitive editing to be properly understood. Finally, too much personalization of a reader or viewer’s newsfeed may dangerously narrow their perspective.
The Canadian Public Relations Society’s White Paper on the future of PR noted the rise in AI was also affecting the PR profession:
“The rapid advance of artificial intelligence (AI) presents a near-term future in which software will increasingly be used to create content; content marketing will be driven by algorithms; bots will manage public enquiries; and decisions of channels and tactics will increasingly be automated, driven in real time by public responses and behaviours.”
For those of us in public relations, the issue is finding new ways of cutting through ever increasing amounts of surface information clutter to cultivate real, constructive relationships between humans.
It’s hard to pitch a story to a bot. If newsrooms are changing, then so is PR.