Google Pushes New Automated Journalism Project in the UK

Can algorithms save local news?
July 10, 2017, 11:00am
Afbeelding: Shutterstock

On Friday I was greeted with the news that my old employer, Baltimore's City Paper, is being shut down later this year by its parent company, the Baltimore Sun Media Group. City Paper is an alt-weekly, a class of aggressively local publications with a bent toward covering culture and news at the fringes (relatively). Most major cities in the US have one like it— sometimes two. Other examples include the Village Voice and Chicago Reader.


Following a succession of closures through a decade of declining advertising revenues, an increasing number of US cities have no alt-weekly at all. Baltimore won't be the first. In a lot of these places, like Baltimore, the internet hasn't really stepped in to fill the gap either, leaving just holes where there should be media coverage and commentary.

I had to think of this as I read about Google's latest push into the world of journalism via its Digital News Initiative, a commitment of $170 million to "to support high-quality journalism through technology and innovation" in Europe. Its latest award goes to the Press Association, (like the British version of the Associated Press). The Press Association will work with a startup called Urb Media on a project called RADAR (Reporters And Data And Robots), the goal of which is to produce 30,000 local news stories each month from open datasets.

The idea of RADAR is that human journalists will identify suitable public datasets that can be monitored for newsworthy events as well as create story templates across different general topics, like crime, health, and employment. The system will then step in to populate these templates with actual news stories with help from Natural Language Generation (NLG) software. It should also be able to auto-generate video and graphics to add to news stories.

The Press Association award will fund five human journalists that will be tasked with building templates, identifying stories, and editing content.

"Skilled human journalists will still be vital in the process, but RADAR allows us to harness artificial intelligence to scale up to a volume of local stories that would be impossible to provide manually," said Press Association Editor-in-Chief Peter Clifton. "At a time when many media outlets are experiencing commercial pressures, RADAR will provide the news ecosystem with a cost-effective way to provide incisive local stories, enabling audiences to hold democratic bodies to account."

RADAR is expected to go live in 2018.

So, will automation ultimately "save" local news outlets? It's all in the scare quotes, I guess. At what point are you just replacing local news coverage with something worse vs. making local news coverage more efficient and driving down costs at perhaps otherwise nonviable organizations? In other words, at what point to readers start to actually care? I'm not super-optimistic about that.