Supercomputers can translate models of complex processes into patterns that humans find meaningful -- and that turns out to be handy. If you've ever seen an exhibit of computer graphics on television or looked at images in magazines, you probably marvelled at examples of scientific visualization (aka "sci-vis") -- striking images rendered in animated, colorful, computer-rendered graphics. Natural phenomena such as storms, or abstractions such as higher-dimensional spaces, can be modeled as glowing colors and flowing shapes on computer screens.
Howard Levine, a former program manager for the National Science Foundation, thinks sci-vis is far more than just a pretty screen: "I think we are seeing the third revolution in scientific thinking tools," claims Levine.
According to Levine, adjunct professor of mathematics at the California College of Arts and Crafts, science grew out of two revolutionary thinking tools: "The ancient Greeks invented deduction -- a logical system of reasoning that started with axioms (statements you could not doubt) and employed precise rules to generate theorems (new knowledge). Early Renaissance thinkers invented induction -- a formal system of rules governing direct observation of and experimentation upon the natural world."
Levine has a bushy gray moustache and a peculiar sense of humor signalled by a sly grin. When I asked him what the third revolutionary thinking tool is called, his grin told me I'd been had: "See-duction, of course," he replied.
"I'm serious about the intellectual revolution," he added, "and I think this tool is particularly important because it brings together the Two Cultures," referring to C.P. Snow's characterization of the opposing ways artists and scientists view the world. "Sci-vis centers are using aesthetic tools from the arts to maximize the scientific information in visual models."
Levine put a videocassette into the VCR and showed me what he was telling me.
First, we watched an animated video of knots. "Bill Thurston is one of the world's best mathematicians," Levine explained, as we observed topology axioms squirming onscreen, "and this video is the proof of his theorem. Thurston's award-winning video, "Not Knot," in fact, preceded the paper proof."
"Computer scientists know how difficult it is to build a computer that can match the visual pattern recognition talents of a human baby who recognizes its mothers poorly-lit face," Levine continued, as we watched supercomputer renderings of human anatomy, roiling purple three-dimensional simulated thunderstorms, psychedelic visualizations of convection currents in the hearts of exploding stars, and models of new drugs engineered in cyberspace rather than wet-labs. "Humans can't do massive calculations, but we know a meaningful pattern when we see one. When you put a computer that is good at calculating and representing the calculations as pictures together with a human who is good at recognizing visual patterns, together they can discover knowledge that a scientist couldn't reach with conventional tools."
The World Wide Web, originally conceived as a way for physicists to publish and view data, is rich in sci-viz resources. In fact, today's high-speed multimedia internetworking technologies were funded by the US government as defense research, in order to maintain superiority in networked supercomputers -- specifically useful for modelling thermonuclear weapons. Like the computer itself, and computer graphics, the child of warfare has matured into an artistic instrument and a new kind of laboratory for groupminds. Sci-vis will grow with the Net. As the raw computing power on every desktop approaches today's supercomputer abilities, over the next ten years, and the amount of number-crunching that can be shunted around a high-bandwidth Net increases dramatically, the Net becomes one big sci-vis lab.
Explore the Sci Vis Home Page at Georgia Tech's Scientific Visualization Laboratory. You can find animations of physical, biological, and chemical models at the Center for Scientific Computing Finland. Or visualize mathematical abstractions with David Banks.
If there is a bright side to our growing partnership with computing machinery it is here, where minds and machines probe the nature of the universe and create beauty together. Is sci-vis a new telescope for the mind? If you are curious about where science and perhaps art is taking us, look for yourself.
Tomorrow home page.