The other day I suggested that maybe selecting which books to publish might just be the only remaining function that we could use to define a publisher.

Now a couple of weeks later, via The Passive Voice, comes news of the impending publication of the first book selected for publication by an algorithm. “Inkitt,* the company responsible for discovering the novel, is an online writing platform where ‘budding authors’ share their work with ‘inquisitive readers’. It relies on an ‘artificially intelligent’ algorithm to bring the two together with the purpose of uncovering ‘blockbuster books’.” Publishing Perspectives also has the story. Algorithm drives selection perhaps, but least the writers are (we assume) humans, though as discussed last year, this isn’t an altogether reliable assumption any longer.

Perhaps our destiny is merely to facilitate the reading of books by machines. These will be books conceptualized by machines, written by machines, edited and produced by machines, and curated by machines. And dare we hope, enjoyed by machines. Relax and enjoy it, proles.


* A matter of boring detail: this is all being presented as much more radical and earth-shattering than it actually seems to be. While Inkitt’s algorithm may be ever so clever, it doesn’t really seem to be doing what’s implied — i.e. reading the books and deciding which one of many to publish. It is bringing together writers and readers and using the readers’ reactions to quantify which book may be a potential blockbuster.

Simpler digital methods of selection have long been enshrined in analog human practice: grip a pin between your digits and plunge it at random into a telephone directory, and lo and behold you’ve located a “character”.