The ability to tell and comprehend stories is a main distinguishing feature of the human mind. It's therefore understandable that in pursuit of a more complete computational account of human intelligence, researchers are trying to teach computers how to tell and understand stories. But should we root for their success?
Creative writing manuals always stress that writing good stories means reading them first—lots of them. Aspiring writers are told to immerse themselves in great stories to gradually develop a deep, not necessarily conscious, sense for how they work. People learn to tell stories by learning the old ways and then—if they have some imagination—making those old ways seem new. It's not hard to envision computers mastering storytelling by a similar process of immersion, assimilation, and recombination—just much, much faster.
To date, practical experiments in computer-generated storytelling aren't that impressive. They are bumbling, boring, soulless. But the human capacity to make and enjoy art evolved from crude beginnings over eons, and the machines will evolve as well—just much, much faster.
Someday robots may take over the world. The dystopian possibilities don't trouble me like the probable rise of art-making machines. Art is, arguably, what most distinguishes humans from the rest of creation. It is the thing that makes us proudest of ourselves. For all of the nastiness of human history, at least we wrote some really good plays and songs and carved some good sculptures. If human beings are no longer needed to make art, then what the hell would we be for?
But why should I be pessimistic? Why would a world with more great art be a worse place to live? Maybe it wouldn't be. But the thought still makes me glum. While I think of myself as a hard-bitten materialist, I must hold out some renegade hope for a dualism of body and spirit. I must hope that cleverly evolving algorithms and brute processing power are not enough—that imaginative art will always be mysterious and magical, or at least so weirdly complex that it can't be mechanically replicated.
Of course machines can out-calculate and out-crunch us. And soon they will all be acing their Turing tests. But who cares. Let them do our grunt work. Let them hang out and chat. But when machines can out-paint or out-compose us—when their stories are more gripping and poignant than ours—there will be no denying that we are, ourselves, just thought machines and art machines, and outdated and inferior models at that.