A caldron of muddy brown water bubbles atop a boiler plate in Professor Robert Birge’s laboratory, lending a faint mad-scientist air to a room crowded with lasers and microscopes.
Once a week, graduate students skim from the top a light-sensitive bacteria native to salt water swamps. The microscopic creature is the guts of a computer memory that may represent the next step toward artificial intelligence, toward a computer that can think like a human.
“Of course, it doesn’t actually work yet,” Birge says casually, almost incidentally. Science-fiction concepts are as appropriate as laws of physics in this chemistry lab at Syracuse University.
Birge practices molecular electronics, essentially, the science of using biological organisms to do the work of transistors and semi-conductors.
It’s a growth area of electronics and one attracting millions in federal and private research dollars. Molecular electronics which combines electronics and biology is seen as the key to such high-tech products as radar-defying “stealth” fighter planes and three dimensional super computers.
Scientists say it may revolutionize electronics as we know it by allowing smaller, faster, more efficient machines.
“Organic circuits will never replace the microchip,” said Birge, 46, one of the nation's foremost experts in molecular electronics. “But within 15 years, I expect molecular electronic systems to be a big chunk of computers and the computer industry.”
That trend is based on the need for smaller and more efficient memory systems as society calls upon computers to perform increasingly complex tasks.
In many cases, Birge said, “natural selection” has already created rugged, versatile organisms ideal for use in electronic circuitry.
“In a sense, we’re admitting nature is smarter than us,” he said. "Speed and density are the issues.”
Most of tomorrow’s high-technology will be based on traditional electronic principles, Birge and other experts say. But the machines and their assignments will demand super-small circuitry; systems that can work thousands of times faster than anything available today.
Birge’s experiments with light sensitive proteins extracted from swamp bacteria illustrates just one reach of molecular electronics. Essentially, he’s seeking to use a microscopic organism known to scientists as bacteriorhodopsin to create a three-dimensional computer memory.
Its microscopic size binary properties lend it to the task. Going 3-D a cube instead of a flat disc drive multiplies potential storage capacity about 10,000 times.
Researchers at Syracuse University’s Center for Molecular Electronics found that a protein memory cube the size of a coffee cup can store 18 gigabytes of data. That’s 18 billion units, or about 4,000 Bibles.
Commercially, molecular electronic computer systems remain years away, Birge said. But other applications of molecular electronics are already being seen, in areas like:
Target Resolution Systems for missiles'. Syracuse University’s Center for Molecular Electronics is working with the Tomahawk Missile System to create sightings with 1,000 times greater resolution than is now available.
Radar Mapping. The Russians are pioneering computer mapping systems, enhanced by a molecular memory system, that can map an area of land fast and on the spot.
Pattern Recognition Systems. The Germans are world leaders in super computer systems that can scan an dread copy in an instant.They are experimenting with a computer tha can read a page of a book in one second.
Radar absorbent materials. The B-1 “stealth” bomber captured public interest because of its radar-defying abilities, but at $1 billion a plane, the bomber is perceived as too expensive to be practical. Molecular scientists are perfecting radar-defying organic coating that can be applied, much like paint, to conventional aircraft.
While the United States pioneered most of the breakthroughs in molecular electronics, it has fallen behind efforts undertaken by German and Japanese scientists and electronics experts, Birge said.
He believes the country will have to resume leadership if it is to be a major player in electronic systems of tomorrow.
“Semi-conductor computers are approaching the limits of what they can do.” Birge said. “New processes are needed if we’re going to reach the next level, artificial intelligence. That’s the goal.”
It may not be as lofty a goal as it sounds. Through organic circuitry, scientists already know they can endow computers with awesome memory capacity. From there, it might only be a short leap to enhance computers with human qualities.
“The sticking point is creativity. How do you give a computer an imagination?” Birge asked. “Some people say we’ll never reach pure artificial intelligence. And I agree. But I think we can get close.”
The solution, Birge said, may lie in allowing computers to mimic the way the brain operates. Creativity is nothing more than mimicry, as the brain associates faces and sights with past memories, he said.
Perhaps a computer can also be taught to “associate” data with reams of information stored in its memory system, and produce a creative response.
Will it work? Birge shrugs his shoulders.
“The potential is absolutely enormous. But we’re not there yet.” He said.
The answer, maybe, lie in a bubbling pot of swamp water.
By Robert L. Smith