Teh vocabularry is the flesh andg rammar is teh bonse

Hot on the heels of the remarkable joke-telling software, that improves the lot of children with communication difficulties through the medium of puns, comes news of a startling breakthrough in the denser field of artificially constructed literature. My chief scientist and one-time co-conspirator Casaubon (surely not his real name) informs me that recent trials of his Automated User-influenced Text Instantiator have borne fruit. Most importantly, does this mean that I might see some return from always having to stand him drinks?

The process is almost childishly easy to explain. To help those in need of narrative stimulation the computer constructs a sort of artificial story. It uses standard English grammar and a selection of vocabulary constrained primarily by consistency of meaning (which I mention briefly below). The main difficulty thus far (and you will appreciate that I only barely understand the underlying technicalities involved in such a vast project) has been in training the computers running the software to appreciate the norms of English idiom and super-grammatical constructions. This particular problem scaled badly with the size of the text, and caused havoc when needing to simulate the plots, characterizations and subtle metaphorical interplay of entire works of fiction. Inculcating meaning in the automata was by comparison trivial, which probably betrays Casaubon’s upbringing among Jesuits. But instead of a good old beating and lots of prayers, the fellow used a subtle mixture of RDF, OWL, PLeX and GANGSTA which, needless to say, has been patented by us both, jointly.

Cas’ tells me that this trial (the first major usage of the software in ira as it were) was conducted with a user cohort of one hundred monkeys. Monkeys? I hear you cry in disbelief. Monkeys, I reply in italics. You see, this group of non-sapient keyboard-hammerers acts (so says Cas: I think he just likes monkeys) as a control, removing much of the otherwise human aspect from the tests.

Even without the human aspect, Cas claims that each “user” has so far constructed an entire novel using only this software and a shedful of bananas. Unfortunately, dear Cas’ went on to admit that the entire corpus of simulated narrative thus far consists of: ninety-six novels almost completely indistinguishable from Neuromancer or each other; three book-length articles not dissimilar from recently accepted submissions to Social Text; and some sort of Collected McSweeney’s. I don’t see Cas’ buying a round any time soon.

Advertisements
This entry was posted in art, literature. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s