‘Patch can create stories tailored for each of their community websites at once, rather than crafting one at a time… forecasts for each of its 1,277 communities can take a few minutes to piece together’
Local news publisher Patch is automating weather reports, real estate sales, and other data-intensive stories, AdAge reports. Patch operates local news sites in 1,277 U.S. communities.
Using the automation tools, Patch editors can prepare templates for data-rich stories instead of the writing the stories themselves.
The result, according to AdAge, is that the automated stories free-up its 110 reporters for larger pieces.
Other news companies using various forms of automated news generation include AP, The Washington Post, Reuters, and Forbes. Urbs Media, in association with the Press Association, uses an AI system for a of local news delivery by constructing templates to localize data contained in national reports. A Swedish company, Mittmedia, is also a pioneer in automated reporting.
OUR TAKE
- This is a example of the ‘force multiplier’ model — a core of 110 reporters able to do more kinds of reporting because an AI system is handling routine news.
- Distinctions between ‘automated’ systems and ‘intelligent’ systems will become more important as the predictive abilities of AI algorithms are used to generate original text.
- Automated systems require structured data using criteria set in advance. They then match it with spots in stories that vary by situation (location, interest, results, etc.), also set in advance..
Patch is using AI to write 3,000 articles a week
AdAge | March 8, 2018 | by Ilyse Liffreing