Last week, the New York Times reported that Narrative Science, a Chicago-based startup, had developed software that would automatically digest sports data and generate a news brief. The software will also create articles out of other material such as a company’s financial reports and housing statistics. Ironically, the software was developed in collaboration with the Medill School of Journalism at Northwestern University.
Narrative Science isn’t the only company in this space. Automated Insights (a company that recently scored $4 million in additional financing and changed its name from StatSheet — could it be because they wanted to use “AI” as their company logo?) has also developed technology that creates “long & short form articles, headlines & summaries written entirely by software, that derive insight from data.”
It’s clear that software like this works best for data-heavy content. And in the aforementioned CJR interview, the founder states that special algorithms for style and tone had to be created; something feasible for data-driven sports and business articles with a consistent tone, but perhaps not that easy for news articles. On Twitter, one user commented that automated content does not equal understanding:
[blackbirdpie url=”http://twitter.com/#!/LawrenceHecht/statuses/116228585798369280″]
Software like this could also be used for mischief. For example, web content farms like eHow could quickly increase the amount of “content” that’s hosted on their sites, solely for the purpose of selling more ads (something they’ve been criticized for previously).
What do you think? Do you think robot-written journalism has a place in the newsroom?
