Google puts news bots on the RADAR

Search giant funds new service in the UK to supply more BOT-generated content for local newspapers and websites 

Purists from the Fourth Estate and readers alike may balk at it but Google has upped the ante for automated journalism with a project to churn out 30,000 BOT-generated news stories per month to kick-start a new service for local dailies in the UK.

The search giant announced on Thursday it has commissioned UK’s Press Association (PA) and data news start-up Urbs Media to put in place the new service, to be called RADAR, short for Reporters And Data And Robots, through a grant of €706,000 (or nearly Rs 5.2 crore), said to be the largest such award to date under Google’s $150 million Digital News Initiative (DNI) Innovation Fund.

“RADAR is intended to meet the increasing demand for consistent, fact-based insights into local communities, for the benefit of established regional media outlets, as well as the growing sector of independent publishers, hyperlocal outlets, and bloggers,” PA said.

“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,” Peter Clifton, Editor-in-Chief at PA added.

PA and Urbs will cull data from publicly available government sources like departments, local authorities, NHS Trusts etc. for AI bots to generate the stories. The funding from Google will help PA and Urbs to set up a team of just five journalists to identify, templatize and edit the data-driven stories. This team will be using Natural Language Generation (NLG) software to produce multiple versions of stories and to scale up the mass localization of news content.

At a count of 30,000 stories a month, or 1000 a day, that too all on topics of local interests, the volume is humongous by any standards. For instance, on any given day the Press Trust of India, our own version of the PA, is understood to put out nearly 2500 stories per day in just two languages (English and Hindi). So clearly, by conventional standards, this small team would be churning out a huge trove of content for publications, print and online, to use.

While the moot question on the impact of AI on Media jobs remains and will be debated for a long time to come, PA’s Clifton insists skilled journalists will still be vital in the process. “RADAR allows us to harness artificial intelligence to scale up to a volume of local stories that would be impossible to provide manually,” he said in a statement released in London.

Projects to use AI and robots in the newsroom are nothing new with several media companies including Associated Press and the Washington Post experimenting with it. For that matter, Urbs already runs a service for newspapers to generate machine written content. In India a lot of work is happening on Natural Language Processing (NLP) and digital information retrieval at institutions like IIT Bombay and IISc, Bangalore to name a few, work at the IIIT Hyderabad is already at an advanced stage particularly where research relevant to the news industry is concerned. Ares of work at the Information Retrieval Labs at the Institute on Structured Information Extraction from the Web and summarization tools is already being implemented in several use cases.  

While RADAR will clearly be of immense value to the mushrooming news websites and “platforms”, as they will be able to access content that is well produced, fact checked and looks good at a low cost, from Google’s perspective it will be a win-win proposition as more content supplied to websites will mean more advertising dollars.

Google announced funding for a total of 107 media projects across Europe on Thursday. It had received over 988 project applications for funding under the Third Round of the DNI Innovation Fund. The trend this year was on fact checking experiments apart from AI-based projects, it said.