you're reading...
Criticism, Fiction, Publishing

The End of Creative Writing as We Know It

I see the end of creative writing on the horizon, and I wonder what it might mean for humanity.

A friend recently forwarded this link about Philip M. Parker, Professor of Marketing at INSEAD Business School, who has labored for the past ten years to develop a computer program that can write books automatically. Yes, you read it right.

keyboardRobotHere’s how it works: A user sets parameters for the program, for example say one wishes to produce a book about narcolepsy (which Parker has done). The program then crawls the web like a search engine, locates and copies relevant information into a database. The information is organized, edited and regurgitated in book form, using an artificial intelligence grammar module, complete with table of contents, notes, bibliography, etc. I’ve seen the result; it’s correct English, and it reads like many other academic texts.[1]

The whole process takes anywhere from ten minutes to a few hours, depending on the complexity of the topic.

Parker has used the program to publish more than 170,000 books. Go to Amazon and search on his name if you don’t believe me.

Currently, the program only produces nonfiction, research type books. But Parker and others see a day in the not too distant future when the program can be configured to write creatively.

Say what? How could a computer replicate the creative process? The whole concept of creativity is centered on an imaginative perspective drawn from a wealth of human experience.

But why not? Why couldn’t someone set the program to write a novel about an inner city youth trying to make a good life, or a love affair amid civil strife in Mali? The program could scour the web in the same way as with nonfiction, compiling reams of information from which it could develop a plot, characters, and then extrapolate an inventive climax and resolution, all in accordance with modern literary theory, which could be coded in by assigning values for aspects of what is considered good fiction. These values could even be varied to the “author’s” or audience’s tastes to produce different genres of work. As for the writing, it would draw on a vocabulary and techniques far greater than any one human writer could possess. And even if it didn’t, frankly, in today’s book market, how good does the writing have to be?

Then we will read; we will be entertained.[2] E-books will include a variety of plot lines and outcomes from which the reader can select depending on her preference or mood.[3] What will be the difference between reading a computer written book and playing a computer designed virtual reality game?

And what of the lowly author? What will be her place in the programmable future? Will we writers always be able to out-think our creations?

Parker also talks about being able to create audio and video programming in the same way, using (in the latter example) computer-generated animation, which is already looking close to reality.

Allow me, non-silicon-based computational unit that I am, to imagine a bit further: At last we will become the things we have made. Sooner or later we will begin to implant computer chips in our bodies, first to regulate our physical well being, but eventually to “improve” ourselves and our minds, creating greater capacity to learn by allowing us to be programmed like computers. We will exist, for a time, as both biological and robotic units. Perhaps the robot half, understanding the many flaws of its biological partner, will wrest control from its weaker side, producing entities of spectacular intelligence that will last for thousands of years.[4] Long before then, however, the programming will condition humans to accept and prefer a specific style of writing, that of the manipulated, computer database-driven text. Creativity in writing will become merely the random generations of the program. Will we still be humans then? Perhaps we are closer to that conclusion than we prefer to believe.

[1] I can’t help imagining this program in the hands of high school and college students—we may never see another original term paper.

[2] And that, comrade, vill be an order!

[3] Some e-books already feature these things.

[4] Starting to sound like a bad “Star Trek” episode, isn’t it?


About Joe Ponepinto

Co-publisher, with Kelly Davio, of Tahoma Literary Review. Author of "Curtain Calls," a featured Kirkus Review. Married to Dona. Dad to Henry, the coffee-drinkin' dog.


30 thoughts on “The End of Creative Writing as We Know It

  1. Skynet has fallen, and the Terminators are among us….but, just THINK of the hours and years save during research….

    Posted by fpdorchak | January 12, 2013, 2:25 PM
    • Yeah, but research is half the fun… for me at least.

      Posted by jpon | January 12, 2013, 10:23 PM
      • True, but for me, with a limited amount of time–and how long it typically takes me to complete a novel (2-3 years–if my series gets accepted by an editor, will have to truncate to a year or less…), I don’t want to spend a lot of time researching. Now, once I can write fiction full time, we’ll see about that. :-]

        No…not a fan of Robo Lit. But in a techno-driven world, you can’t keep creativity down. This remind me of those No-Frills Books, by Jove Books, that came on in the early 80s. I bought one for “MYSTERY,” ISBN 0-515-06248-0, and it was 58 pages. I no longer remember how good it was, but it was, literally, a story with no frills. :-]

        Joe, I wouldn’t worry about Robo Lit. You know, unless it starts kicking our collective asses in the critical reviews…. ;-]

        Posted by fpdorchak | January 12, 2013, 11:40 PM
  2. Please, not so much doom and gloom on a warmish Saturday morning! Tell me it ain’t so!

    Posted by shadowoperator | January 12, 2013, 3:07 PM
  3. Sounds like a premise for a good Sci-Fi novel.

    Posted by Brian B. King | January 12, 2013, 3:20 PM
  4. Joe, Joe, Joe, you’re a worrier. I mean, you’re doubtless right about the direction of robo-lit. We’ve been heading that way since the dawn of computing, so it can be no surprise that we’re nearing the goal of human-less creativity. The way things are going, computers will probably need to develop a kind of reverse Touring Test. Humans will be required to interact with computers in order to prove their fallibility, and thus their ownership of the device. (Beginning to sound like a Twilight Zone episode.)

    But to the point of writing. Sure, they will eventually write perfectly crafted, grammatically correct stories. But few will read them. We want out novels to be imprecise, asymetrical and rough around the edges. We want blurry images, impressionistic and approximate, things abhorent to the computer. Computer writing will be photographic. Human writing is paint brush. Many years ago, a friend of mine lamented that representitive visual art died with Matthew Brady and the photographic image. But that hasn’t happened. And it won’t. And neither will a take over by scribe-bots.

    No computer could ever write as badly (or as popularly) as Fifty Shades of Gray. Nor will it print out the depth of Steinbeck, the quirk of Lori Moore or the smirk of Vonnegut. The worst and best of human writing will continue. The middle may belong to the silicone mind, but the extremes will always be human.

    The Enterprise may make Warp Factor nine, but it is the people within who boldly go.

    Posted by Jon Zech | January 12, 2013, 3:32 PM
  5. Blehhhhhhhhhhh. Sounds awful.

    Posted by Cassie | January 12, 2013, 4:17 PM
  6. Sounds like you’re ready to research posthumanism and Singularitarianism (what Cory Doctorow and Charles Stross have humorously called “The Rapture of the Nerds”). Ray Kurzweil would be a good place to start. I heard he just got hired by Google.

    Computer-driven creative writing started as early as 1984. Seriously: http://www.ubu.com/historical/racter/index.html (But, as would be expected from such an antediluvian computer, “The Policeman’s Beard is Half-Constructed” works better for fans of surrealism or absurdist poetry than most of today’s fiction consumers.)

    Posted by akhoffman | January 12, 2013, 4:32 PM
    • Actually, Radical Evolution by Joel Garreau is probably a better intro to trans- and posthumanism because it presents both the optimistic and pessimistic angles on it. Already getting a bit dated, in light of things like Google Glass, self-driving cars, Siri, brain-to-prosthetic-limb interfaces, etc., but still covers some useful territory. (I read it as part of my research for a spectacularly shitty first novel.)

      Posted by akhoffman | January 12, 2013, 4:41 PM
    • Reminds me of the exposé a group of writers pulled on Publish America. To prove that PA would “publish” anything, no matter how poorly written, they had a computer compose a novel. It was mostly gibberish, but PA snapped it up just the same. I tried to find a link but came up empty (only checked one page tho).

      Posted by jpon | January 12, 2013, 10:41 PM
  7. Nah. What comes around goes around, and we humans are nothing if not restless. Even if the Star Trek episode were brought to life and computers began to write our fiction, we’d quickly get bored with the sound of it and return to the organic model as a novelty. A computer can’t construct an original metaphor, after all, or expound on a theme. They only know what you tell them—and even a human writer doesn’t know what (s)he’s trying to say until it’s written down.

    Posted by Averil Dean | January 12, 2013, 4:46 PM
    • I agree with you, Averil, and I hope you’re right. But the next time I try to talk to a computerized system over the phone as though it’s a real person, I may change my mind.

      Posted by jpon | January 12, 2013, 10:46 PM
  8. My father traveled his whole life. He never owned a home. You know what he did? He sold encyclopedias door to door to military families on bases around the world. He died in ’93. His job no longer exists.

    Online books have replaced encyclopedia salesmen. Technology has wiped out numerous careers. Why not creative writers?

    My answer may be naive, but I’m not convinced a computer program can replicate the nuances of creating quirky characters. And I don’t think a program is capable of exploring the depths of discovering what people are capable of doing when we toss bombs at them. Without a human perspective, that would be lost.

    But if it is possible and does happen, we’ll have to adapt.

    Posted by Darrelyn Saloom | January 12, 2013, 4:54 PM
    • Right now computers can’t compete creatively with humans, but that may change. I only hope that our adaptation, should computer generated lit become the standard, is that we find challenges rather than acquiescence.

      Posted by jpon | January 12, 2013, 10:49 PM
  9. Fascinating post! Towards the end, you write, “Long before then, however, the programming will condition humans to accept and prefer a specific style of writing, that of the manipulated, computer database-driven text.” I’d also add that we’re pretty much there already (as you suggest in your final sentence) insofar as we’re already more or less conditioned by the media to consume whatever the mainstream mass-marketplace produces. Adding computer-generated literature to the mix will certainly appeal to our increasing thirst for immediate gratification.

    Posted by Marc Schuster | January 12, 2013, 7:25 PM
    • I’ve always been fascinated by what people will accept if they trust the authority that delivers it. For example, I’m watching the NFL playoffs while I respond to comments, and just saw a commercial for the movie “Hansel & Gretel: Witch Hunters.”

      Posted by jpon | January 12, 2013, 11:00 PM
  10. …and come to think of it, certain genre’ publishers already have writing “machines” that do little more than plug in a formula, add a few tropes, and grind with standard, uniform dialog.

    Posted by Jon Zech | January 12, 2013, 8:51 PM
  11. oy gevalt. a fantastic sf story by Walter Miller tells of an out-of work actor, turned janitor, who ends up getting on stage with the professional robot/puppet actors who’ve taken over the business and things get wild. i think we’re always fighting these big machines as Alex of A clockwork orange puts it….:http://en.wikipedia.org/wiki/The_Darfsteller

    Posted by Stephanie Barbé Hammer | January 14, 2013, 5:42 AM
    • Yup. There’s always someone who thinks another machine is the answer. But what I find interesting is that every piece of technology, even the most well intentioned, always brings with it a new set of problems that wasn’t foreseen. In the long run we are usually better off, but perfection may not be possible… continued in next comment

      Posted by jpon | January 14, 2013, 8:57 PM
  12. I just don’t believe it could be good or that I’d want to read it, except once to see what it looked like. However, I’m thinking that this would make a fabulous plot for a sci fi novel. Write it in a literary style and I’d read it. Who’s going to grab it first?

    Posted by girl in the hat | January 14, 2013, 6:39 PM
    • You may have to get in line. But continuing my thought from the comment above, it would be interesting to see what problems, aside from bad prose, such a technology might cause. Eventually, of course, the machines would tire of writing stories about people and begin to write them about themselves instead. Gone will be the human struggles for love and accomplishment, replaced by the search for the perfect code.

      Posted by jpon | January 14, 2013, 9:02 PM


  1. Pingback: The End of Creative Writing as We Know It | Poetry - Spare | Scoop.it - January 12, 2013

Tahoma Literary Review Now Open for Submissions

TLR is officially open for submissions of poetry, fiction, and nonfiction. To find out more about this new (paying) literary journal, please visit us at Tahoma Literary Review.

Enter your email address to subscribe to Joe's blog and receive notifications of new posts by email.

Join 7,385 other followers