Preface: This is my first ever blog and my first ever blog post. I did it at the urging of my friend, who recently started his own blog and, I guess, wanted some online company. But I also have wanted to start blogging for a long time, just because it seems like such a quick and easy way to express my opinions on a regular basis, and develop my writing. The second reason is actually a pretty pressing need, as I write for Nazar, an online magazine at The University of Texas at Austin. I use the term “write” loosely, however, as I haven’t really written an article on there in 6 months…
So, you might ask, why am I wasting my time writing for this blog in the corner of the web when I can write for an audience of over a thousand visitors a month at Nazar? Well, it’s basically because I just haven’t been able to write anything for Nazar. Or rather, I haven’t been able to write anything good. And it’s not because I can’t write anything good (I hope); no, it’s because I just haven’t written anything longer than about half a page in so long. I’m out of writing shape, and I need to get in some reps, get in some practice. So you, lucky reader, will get to read the pieces I whip out as I practice honing my technique and get back in the flow of writing. I hope to write one post per day, Monday through Thursday, so keep checking back for fresh content and I hope you enjoy my first post below!
Ever since I’ve been a little boy, I’ve wanted to be a professional novelist once I grew up, to make a career out of writing. Upon hearing this, my mom would always become very upset. A traditional Indian mom, she of course pushed me towards Engineering and Medicine. Writing, she would say scornfully, is fine if you want to be working in Burger King on the side. I always thought this was really unfair. I mean, plenty of people make a living off of their writing, and a few even strike it rich.
I resolutely held my stance until I came across something that both fascinated and alarmed me: a computer program that could write news articles. Narrative Science, a Chicago-based startup founded in 2010, had brought together programmers and journalists to develop an artificial intelligence writing platform named Quill. They worked in close conjunction; the journalists built sets of templates to describe parts of the reporting spectra and the programmers worked to gather large volumes of data about each field, so that the program could gain a “context” of sorts. Then, given a set of information concerning an event (for example, the box score of a baseball game), Quill will churn out a news article describing it.
What’s most surprising is that the resulting pieces don’t at all seem computer generated. In fact, the algorithm can write better than some people! Check out this short article. If you missed its by-line, it would be incredibly difficult to tell that this wasn’t written by a living, breathing human being.
Right now, the scope of Narrative Science’s pet program is still limited to sports and financial news articles, fields that allow large amounts of numbers to be gathered and make the program’s life a little easier. Right now, by taking on small jobs like reporting on children’s baseball games, Narrative Science is actually expanding the boundaries of writing and not really infringing on any writer’s turf.
But all that will change, and soon. There are already multiple competitors to Narrative Science, and the competition will only grow as people become more aware of this technology and the success Narrative Science is having with it. In addition, the very technology will improve and expand, and if its sophistication progresses at anything close to Moore’s Law, we’ll be seeing robot Shakespeares in no time. What’s more likely, though, is that we’re still quite a ways away from that happening, but it’s still very possible that computers will be dominating writing in the near future. In fact, Narrative Science’s CEO, Stuart Frankel, sees over 90% of news articles being written by computers 15 years from now.
Other avenues of writing, such as creative writing, opinions and editorials, etc. will be a bit more difficult for computers to penetrate, but I have no doubt that it’ll happen eventually. And that’s why I’ve been rethinking my desire to actually become a writer. How would I be able to compete with a writer who has perfect knowledge of every book, poem, or story every written, a writer who could process all possible permutations of word choices and sentence structure and find the optimal combination? My hypothetical future job would be computer-sourced! But this problem isn’t specific to writers, I’ve realized. Computers are already replacing cashiers at some McDonalds and are poised to encroached onto nearly every field of work. If writing, one of our most undeniably human endeavors, can be effectively “solved” by an algorithm, what can’t?