TThe visible properties of this piece are easily and briefly enumerated.
This is a portrait of the great Argentinian writer Jorge Luis Borges. It was created out of images gathered from Google using the words to his short story “Pierre Menard, author del Quijote” as search terms.
Ten images were selected at random for each individual word in the story, for a total of 11,631 images. Here are some of the images, with their corresponding terms.
Borges' story is available for free here, and I encourage anyone to read it; but here's a summary for the impatiently uninitiated:
The entire piece is written in a tongue-in-cheek manner (Menard is the author of a "list of verses which owe their efficacy to their punctuation"), and we are meant to laugh at it. At the same time, there is something very familiar about what Menard is attempting to do. It has to do with the Anxiety of Influence, of course, but also with the process of reading. In Borges' words:
He did not want to compose another Quixote—which is easy—but the Quixote itself. Needless to say, he never contemplated a mechanical transcription of the original; he did not propose to copy it. His admirable intention was to produce a few pages which would coincide—word for word and line for line—with those of Miguel de Cervantes.
Mendard’s bold ambition could only be fulfilled by completely become Cervantes (that is, conveniently forgetting the 300 years of history that lay between them), or else by becoming Quixote himself.
This is an image made using the same process. The picture is of an aging Borges, wearing a pot on his head and generally being a goof. You can click it for a full-sized image.
The reason I'm bringing up Borges, here, and Pierre Menard specifically, is that the method employed to produce these images is based on Menard's fictional attempt to re-create the Quixote. The images were drawn using OpenGL shaders. Shaders exploit the processing power of modern graphics cards to perform loads of calculations much faster than possible using a computer's preoccupied CPU.
To eschew excessive technicality: the images were produced using a program that draws random images in random places. After each iteration, it compares the image it has just produced with the "ideal" image it wishes do create (its Don Quixote, if you will). If the drawn image has grown closer to the ideal, the iteration is kept; if not, it is thrown away.
Ordinarily, this would be a horribly inefficient way to draw an image. In fact, one can never possibly create the original image — only draw ever-so-slightly closer.
In some ways, it resembles the old hypothesis of infinite monkeys rewriting Shakespeare on infinite little monkey typewriters. This is what I find so poetic about this whole exercise: the sensation of stumbling haplessly toward infinity, knowing full well that I will never get there, because I have yet to evolve further than a monkey.
Pierre Menard phrases it much better than I can:
“My undertaking is not difficult, essentially,” I read in another part of his letter. “I should only have to be immortal to carry it out.”
And so I will leave you with this, a portrait of William Shakespeare, created by my little army of GPU monkeys using fan-drawn "furry hentai" art. Like love or politics, it's completely safe for work unless you over-scrutinize.
Note: you can find source code and instructions for building your own Pierre Menard on my GitHub.