MIAMI — In a high school science class, one quickly discovers that the eyes, so often taken for granted, do an awful lot of work: it’s a decision of the brain to both obscure the nose from our line of sight and upright the images on our retina, which initially appear upside-down. Upon learning this, I felt, at once, that my brain was miraculous and that the whole process was completely tenuous. We later learned about periphery, colorblindness, and “schemas,” patterns of thought that organize groups of information, and how rapidly such schemata become subjective. Vision, it became clear, is just one component of perception, which itself is a many-armed beast.
SEEING: What Are You Looking At?, which explores how brains, eyes, and artificial intelligence perceive the world, is the first special exhibition at the newly reopened Phillip and Patricia Frost Museum of Science and the second incarnation of the exhibit itself — this is its US debut, having traveled from Ireland’s Science Gallery Dublin. Curated by several people — Kate Coleman, an oculoplastic surgeon and founder of Right to Sight; Lynn Scarff, the director of Trinity College Dublin’s Science Gallery; Gerry Lacey, an associate professor of Community Science at Trinity College Dublin; and Semir Zeki, a professor of neuroaesthetics at University College London — SEEING examines their unique concerns: vision, larger ideas about perception, the democratization of eye care, and how all of this might apply to A.I., whose sense of sight is still even more faulty and subjective than our own. Most of the work included is by interdisciplinary artists, along with a few educators in the field of design and vision.
Frost Creative Director Alexandra Kuechenberg has curated a sprawling exhibit spread over two floors of the museum. Many of the works are visually pleasing in a way that might appeal to children: Kurt Laurenz Theinert’s “Magical Colour Space” is a room comprised of striped wallpaper and a light that slowly shifts its proportions of red, blue, and green, until the colored stripes seem to glimmer and disappear. The dizzying effect satisfied my gleeful inner child, and this is fine; in the most basic sense of the word, optical illusions are always stimulating, simply because they always work (hence the popularity of Magic Eye books among grade-school kids).
Most of SEEING, however, vacillates between critical examinations of A.I.’s visual capabilities and experiential meditations on the fragility of sight that help to foster empathy for those who are vision-impaired. In “Sight Without Light: Exploring Human Echolocation,” a piece by Story Inc. and Daniel Kish (president of World Access for the Blind) visitors are invited to experience the world the way Kish does. Kish, whose eyes were surgically removed as an infant, has taught himself to navigate the world through echolocation, clicking his tongue and allowing the sound to bounce off objects around him; at the museum, guests are blindfolded and instructed to make a loud ssshh noise as they cup their ears, listening as the sound changes when an obstacle (a small ball, for example) approaches.
This interactivity is prevalent throughout. An ultrasound cane is on display, and an autorefractor available for off-the-cuff eye examinations. It’s the exploration of A.I., though, that is SEEING’s crux, as well as the question: Can A.I. develop its own kind of empathy? Here, SEEING starts to become less visual exhibition, more educational laboratory. “The Innovation Engine,” by Frederik De Wilde, Jeff Clune, and Anh Nguyen, allows you to step in front of a computer screen while it runs through its database, comparing you to its stored images, figuring out how it might label you. On my tour of the exhibit, it determined Kuechenberg was chainmail, due to the finely dotted pattern on her shirt. (It thought I was a beige coat, which I attribute to absolutely nothing.) If your expectations are high, crude A.I. is consistently disappointing, glitching in a way that’s often comical.
“20/X,” a piece by interdisciplinary duo McMullen_Winkler, enables you to see objects as a computer might — underneath a white fabric bell, where the image of the object is projected and self-adjusted. Toying with the toggles on the screen, I watched a bouquet of flowers shift into multicolored, pixelated simulacrums. I wished, instead, to look at a flower, soft and unencumbered, but that is not the point — I was watching a computer learn flowers, their droopy edges and round shapes. Advancing technology for the purposes of research and identification feel significant, but perhaps this research can be applied to assisting the visually impaired, too — if an eyeless mechanism can learn to detect the intricacies of something organic, surely the blind can eventually maneuver through a world that’s more accessible?
In another vein, perhaps an impending singularity isn’t so freaky, if we all might harmoniously coexist with robots. My favorite piece was “3RNP,” designed by Patrick Tresset, where three robots named Paul draw a human model. The Pauls, which seem to be made entirely of gears and webcams, are seated at school desks; volunteer models inevitably look awkward as the robots’ arms — which look like protractors — piece together scratchy portraits. The finished drawings are hung on a wall nearby, in rows of abstract, pen-etched people. While there’s something unpolished and silly about this, its potential implications for the future make the process seem almost quaint. Strangely enough, if you look closely, the details of the portraits’ expressions are surprisingly detailed — almost human.
SEEING: What Are You Looking At? continues at the Phillip and Patricia Frost Museum of Science (1101 Biscayne Blvd, Miami,) through October 20.
The post An Interdisciplinary Look at the Differences Between Human and Artificial Sight appeared first on Hyperallergic.