In an article today in The Daily Dot, I talked with writer Aaron Sankin about the proliferation of automated content in news stories. The article was inspired by a couple recent events.
Two weeks ago, Karstad University’s Dr. Christer Clerwall published the findings of his study of automated content vs. human-written content in the 2014 volume of Journalism Practice. Dr. Clerwall used samples from Automated Insights’ StatSheet product (which Robbie Allen and I built over three years ago) to compare against actual journalist-written articles of the same NFL game.
His conclusion: Not only is automated content virtually indistinguishable from human-generated content, but in most cases it is deemed as more credible and trustworthy.
By the way, I’ve since spoken to Clerwall, and it’s obvious he gets where automated content is going and why it’s so valuable.
As you might imagine, this news got picked up by a lot of outlets, including PandoDaily and the Daily Dot.
Then, the Monday, March 17th Los Angeles earthquake story was “broken” in the LA Times with a computer-generated story published by QuakeBot, which monitors and reports on the data feed from the U.S. Geological Survey.
People were fascinated, and rightly so. This is a perfect example of machines being able to break news at a moment’s notice. However, much like those automated weather service warnings you get on your television and radio, it’s not exactly brand new.
In fact, you’ve probably already read your share of automated content, especially in the areas of sports, finance, weather, real estate, and personal fitness.
Since Sankin had already written about Automated Insights’ content being used in the Clerwall study, he got in touch with me to talk about the past, present, and future of automated news.
He makes some great points:
“The initial reaction among many journalists was a combination of ‘woah, computers can write new stories?!’ and ‘uh oh, robots are coming for our jobs!'”
This is absolutely true. It still happens every now and then, when journalists and other writers cast a jaundiced eye at automated content. But as I’ve said from the beginning, we’re not in this to replace writers (especially good writers). We’re in this to produce content where human writers can’t write (i.e. Yahoo’s Fantasy Football Recaps) or don’t want to write — in cases where we produce instant articles that can then be updated, polished, or elaborated upon by human writers.
My favorite part is at the end, when Sankin concludes, after posting some examples of Automated Insights content:
“If you randomly stumbled across any of these pieces of content, could you tell they were created by a computer? The larger question probably is: would you care?”
That last sentence is important. We’re already working with customers who not only don’t mind publishing our automated content, but some are actually promoting the fact that it was created by a machine.
This technology isn’t going away, but it also isn’t going to put solid journalists out of a job. Not soon, anyway — with technology, you can never say never.