damn there's some smart people on here! Had me lost though, i think i need to spend some more time on wikipedia,lol.
found this on chaos theory and cpu, waaay back from 1999:
Order from Disorder
Researchers from many areas of study look to chaos theory for progress:
www.business2.com...
form here onwards the article gets juicy:
Inside a small laboratory at the Georgia Institute of Technology in Atlanta, a revolutionary method of computing floats in a small plastic dish
underneath a microscope's lens. The brain of this computer is quite literally that — not silicon but brain matter itself: neurons, the basic
building blocks of the brain, dissected from leeches.
Digital computers that work by turning bits on or off don't make the grade when it comes to "creative" tasks like handwriting or speech
recognition. What's needed are computers "that are more like human brains and can rewire themselves" to find solutions, says William Ditto, head of
Georgia Tech's Applied Chaos Lab.
"We're hoping within five years to have a chip with living matter on it to solve specialty problems in ways that conventional computers can't,"
Ditto says. "These problems range from pattern recognition to simulating extremely complex real-world physical systems like economic and
environmental models." (See "Big Ideas from Small Creatures," Jan. '99, p100.)
Still a technological infant, the biocomputer consists of just two leech neurons, chosen because their relatively large size makes them simpler to
wire. Eventually, the biological neurons may even be replaced with silicon substitutes, Ditto points out. Completely organic or not, though, in this
early incarnation, the wetware computer currently isn't doing anything a 4-year-old can't. The impressive thing is that it's doing it in much the
same way that a 4-year-old does.
Once the leech neurons are interfaced via electrodes to a PC, each one can be programmed to represent a specific number. The PC then "tunes" the
neurons to undertake a specific task. Addition is its only trick thus far. After the PC stimulates them, the cells link with one another, talking to
each other very much like the neurons do in a human brain. Within several seconds, the neurons come to a solution. Not a groundbreaking result, but
it's the brain-like way in which the leech computer tackles a problem that may revolutionize data processing, Ditto explains.
"A Pentium computer does really dumb calculations very quickly," he says. "But a 4-year-old can come to decisions based on partial information and
figure out how to do additional operations."
The key to making a computer more brainlike, explains Ditto, is that "each element can't behave randomly or you'd just get gibberish. You need the
elements, neurons in this case, to behave chaotically so they can interact with each other adaptively."
And that's an entirely new direction for computing, veering off the traditional path that "faster is better." According to Ditto, the key is not to
increase the speed of the processor but to increase the number and flexibility of connections between its elements. Then, he says, we'll be closer to
mimicking the chaotic complexity of our own biocomputers.
"Living brain cells are like nature's transistors," he says. "We're not really smart enough to know what they're doing, but we can still use
them."