Nar­ra­tive Sci­ence is at the fore­front of com­put­er automa­tion.  At Media Jobs we have brought you numer­ous sto­ries dis­cussing how data visu­al­iza­tion is all the rage right now and will be into the fore­see­able future, and for good rea­son. The vast amount of data being gen­er­at­ed and col­lect­ed every sec­ond of every day on every pos­si­ble top­ic is almost use­less with­out a good way to ana­lyze it and present it in an under­stand­able way. And they were all writ­ten by humans. But in cer­tain sit­u­a­tions, those charts, graphs, and oth­er unique visu­al­iza­tion tools still aren’t the best way to trans­fer infor­ma­tion to peo­ple. Some­times a con­cise and infor­ma­tion-packed sto­ry is the most pleas­ing and under­stand­able way to absorb infor­ma­tion, and that’s where Nar­ra­tive Sci­ence excels, no humans required.

Tell Me About The Narrative Science Game

If you ask Google or Siri or Cor­tana this ques­tion (which is actu­al­ly a state­ment, I guess), you’ll prob­a­bly get a final score — assum­ing that you spec­i­fy the game you’re refer­ring to, of course. But soft­ware devel­oped by Chica­go-based Nar­ra­tive Sci­ence does­n’t wait for you to ask. Instead, it col­lects the rel­e­vant infor­ma­tion about the game and writes a sum­ma­ry in para­graphs we’re used to read­ing from sports writer sum­maries, using a tech­nol­o­gy known as NLG, or nat­ur­al lan­guage gen­er­a­tion. Not just sports games, though — any sto­ry that can be told from col­lect­ed data.

In fact, one of Nar­ra­tive Sci­ence’s largest cus­tomers (if not the largest) is Forbes, the dis­tin­guished finan­cial pub­li­ca­tion and web­site. They use Quill, the afore­men­tioned soft­ware that Nar­ra­tive Sci­ence pro­vides, to write earn­ings sum­maries for com­pa­nies on a reg­u­lar basis. The soft­ware is even list­ed as one of the con­tribut­ing authors in that sec­tion of the web­site. It also adapts the lan­guage used in sim­i­lar arti­cles writ­ten by real live humans, so if you’re only want­i­ng the quick run­down on a ball game or a stock­’s earn­ings infor­ma­tion, you’d nev­er know that the para­graph you just read had no human hands involved.

News Written by Robots For Robots

In the two exam­ples already men­tioned, and many oth­ers for that mat­ter, this non-human sto­ry­telling makes noth­ing but sense. Why pay a human writer to essen­tial­ly dis­sem­i­nate some num­bers by throw­ing them into a few unin­spir­ing sen­tences? It’s a cost saver, and it gets the point across with just enough nar­ra­tive to be inter­est­ing — or more inter­est­ing than a sim­ple score or set of num­bers any­way. But there are also con­cerns about where this trend might lead in the future as the soft­ware advances.

Nat­u­ral­ly and right­ly so, writ­ers have the most con­cerns about this. But a more in-depth analy­sis of what’s hap­pen­ing now and a like­ly sce­nario of the evo­lu­tion in NLG writ­ing shows how infor­ma­tion could become so per­son­al­ized that it actu­al­ly becomes more harm­ful than help­ful to intel­lec­tu­al pur­suits. I won’t get into all that here, but this Slate arti­cle does a good job of express­ing those concerns.

All Skynet-ish con­cerns aside, the tech­nol­o­gy that Nar­ra­tive Sci­ence has devel­oped is a step for­ward in the AI field and could lead to any num­ber of inter­est­ing and lucra­tive devel­op­ments down the road, which is why we think it’s a com­pa­ny that shows a lot of promise and would be a great career opportunity.