In Futurama, a cartoon set 1000 years in the future, robots are common sights, performing a multitude of tasks, while also providing a lot of the series’s humor due to the vast array of seemingly unrobotic jobs they perform. A great example is “Humorbot 5.0,” a stand-up comedian/late-night talk show host. The concept of a robot knowing anything about ‘being funny’ leads to comedy within the show as he provides snappy jokes like, “Super-collider? I just met her… And then they built the super collider.” Although this takes place after the year 3000, reality may be catching up in the form of a technology known as Narrative Science.
Developed at Northwestern in a joint effort between the Engineering and Journalism departments, Narrative Science combines editorial expertise with regimented, computer algorithms to create clear, consistent narrative reporting of data, allowing readers and data experts to gain greater insights. Companies like Forbes use Narrative Science to generate online articles reporting on corporate earnings statements. Narrative Science takes the hard data (stocks, IPO, pricing, etc.) and automatically presents it not in a graph or chart we expect from computers, but in a readable, concise narrative story one would expect from a trained journalist. In terms of publishing, Narrative Science currently gets the most use in data-heavy fields, like finance and sports, but the technology could eventually breach into all types of reporting. Already, it has generated news stories from “unstructured data” like Twitter to ‘write’ a story on Newt Gingrich’s Twitter feed.
Granted, none of this reporting is likely to win the Pulitzer Prize (until that, too, is awarded by robots!). But it is looking like this will become more and more common. On the publishers side, it is simply a matter of economics. Keeping a full-staff of writers to scour data, write a story based on it, edit (humans make mistakes, of course), and publish is both expensive and time-consuming, something Narrative Science can do on its own in less time and for less money. This is all especially true for data-heavy reportage like financial information and sports statistics, where ‘analysis’ isn’t as necessary as simply ‘data-presenting.’ However, computers are analyzing machines, and if they are used to do the types of trading Narrative Science reports on, what is to stop these robo-journalists from analyzing their data as well?
There would seem to be a line that robots cannot cross. “Humorbot 5.0″ is funny, but for all the wrong reasons, so I can’t see robots replacing fiction or more ‘thought-based’ writers. (I sure hope not, seeing as that’s basically what I’m doing here!) Rather, robots are meant to handle increased levels of information and data, as they always have been. The real question, however, is if this proliferation of robotic writing and creation will lead to a general cultural acceptance of robotic-generated material? If society becomes more used to the presence of robots as cultural creators/curators, their expansion into other fields seems natural. Perhaps it won’t be long before we see “Humorbot 5.0″ following Leno on NBC.