Our Robot Writers Will Be Your Valued Colleagues

by James Kotecki

The spokesperson for a company that automates news articles says that software will collaborate, not compete, with journalists and other workers.

At Automated Insights, we’re often asked if Wordsmith, our “robot writing” platform, threatens human jobs. It doesn’t, and we’ve written in detail about why. Last year, Wordsmith automatically turned raw data into one billion human-sounding narratives without displacing anyone.

Beyond practical concerns about lost income, the jobs question hints at a deeper uneasiness. By doing tasks that were previously exclusive to humans, is automation making the world a less human-friendly place? Does the mechanization and computerization of complex tasks chip away at our human identity? Are we building a world that is more efficient but less fun, a world in which people are just the messiest components of machine-driven processes?

We can’t speak for all forms of automation, and to be sure, existential questions about employment and economics deserve a debate that goes beyond the scope of any one business. But we do believe that our company has a place in that debate. At Automated Insights, we make jobs, data, and user experiences more human, not less.

Humanizing Jobs

In our experience, automation doesn’t make people feel like robots. Quite the opposite, in fact. People feel like robots when they have to do work that could be automated.

The writing of corporate earnings stories is a clear example. The Associated Press used to take valuable reporting time to do what Vice President Lou Ferrara calls “data processing” — the mind-numbing, nearly formulaic task of pulling numbers from quarterly reports, performing the necessary calculations, and writing a recap in AP style.

Needless to say, reporters didn’t like doing this rather mechanical part of their job. Now that our Wordsmith platform automatically generates those narratives, journalists have more time to analyze a company’s business model, explore industry trends, or contextualize a quote from the CEO. In fact, reporters often work with Wordsmith, adding their personal touch to stories initially drafted by machine. Our automation technology did not take any jobs at The Associated Press — it made the newsroom a more human place to work.

Humanizing Data

Wordsmith also brings a human touch to the rapidly growing, economy-spanning world of Big Data. By finding insights in business data and presenting them in plain language prose, Wordsmith turns overwhelming amounts of information into actionable business intelligence.

At a Fortune 100 company, for instance, the platform turns data on team member performance into reports that give customized advice to workers and their managers. It uses the language, style, and tone that is most comfortable to its intended audience — in this case, that of a business analyst. Instead of wading through pages of inscrutable graphs, Wordsmith helps readers easily find what they need to know.

Thus, automation turns raw data into real-world impact. There simply aren’t enough human data scientists to do this work, especially when it needs to happen in real time. Our white paper, “Big Data Needs Big Insights: The Business Case for Natural Language Generation,” explores this subject in greater detail. Overall, Wordsmith delivers human-quality analysis that wouldn’t have existed otherwise.

Humanizing Experiences

Our platform helps businesses relate to their customers as people, not just numbers. By automating personalized stories for readers based on their specific data, Wordsmith makes user experiences more fun and more engaging.

For example, Wordsmith recaps millions of Yahoo Fantasy Football matchups every week of the season. It’s as if a snarky sports reporter is writing about each user’s specific fantasy football experience and including clever jokes about less-than-stellar performances. Wordsmith turns pure statistics into a compelling, human story.

In the realm of personal health, Wordsmith turns data from GreatCall about the location and well-being of older Americans into narrative reports that give their adult children peace of mind.

“Presenting user data as a chart or graph just didn’t feel like the right way to represent people’s lives,” says Krijn van der Raadt, GreatCall’s Vice President of IT and Software Development. “Our customers agreed — a narrative summary of somebody’s week is a much more natural fit.”

Wordsmith adds a human layer onto the data and makes it personal.

Beyond “Man vs. Machine”

NPR’s Planet Money recently put our technology to the test. In a man vs. machine race meant to mirror John Henry’s railroad duel with a steam drill, Wordsmith squared off against a human journalist to recap a story about Denny’s quarterly earnings. In the end, reporter Scott Horsley wrote a more colorful story — but Wordsmith worked much faster. Host Steve Henn declared Wordsmith the winner, and Horsley congratulated “our future computer overlord, Automated Insights.”

But here’s the truth: Wordsmith isn’t in locked in a battle of people vs. machines. We don’t think about it that way. Our paradigm is people plus machines. Humans are integral to every step of our process, from building and configuring Wordsmith to adding context to its narratives to engaging with the content it generates.

For us, the purpose of automation is to make jobs, businesses, and user experiences better for the people involved. In our data-saturated world, people can feel alienated by or beholden to machines. Automated Insights makes the world a better place to be human.

James Kotecki is Automated Insights’ Manager of Media & Public Relations.

Learn more at AutomatedInsights.com and follow @AInsights on Twitter.

Have your own opinion? You can leave a response below.