Posted on 07/14/2004 12:55:29 PM PDT by LibWhacker
A team of computer scientists and mathematicians at Palo Alto, CA-based Artificial Development are developing software to simulate the human brain's cortex and peripheral systems.
As a first step along the way, the company recently disclosed that it has completed the development a realistic representation of the workflow of a functioning human cortex. Dubbed the CCortex-based Autonomous Cognitive Model ('ACM'), the software may have immediate applications for data mining, network security, search engine technologies and natural language processing.
The first ACM computer 'persona,' named 'Kjell' in homage to AI pioneer Alan Turing, was activated last month and is in early testing stages. CCortex, Artificial Development's high-performance, parallel supercomputer, runs the persona simulation.
Marcos Guillen, president and CEO of Artificial Development, disclosed the ACM's emerging cognitive capabilities at a recent Workshop on Cognitive Systems, co-hosted by Sandia National Laboratories and the University of New Mexico.
Guillen commented that the ACM is intended as a test-bed for future models and is still incomplete. While the Kjell persona uses a realistic frontal cortex, motor and somatosensory areas, it still lacks the visual and auditory cortex areas, two of the most important cortical structures. Other structures, such as the hippocampus, basal ganglia and thalamic systems, are still being developed and are unable to perform most normal functions.
The ACM interacts with trainers using a text console, reading trainer's input and writing answers back, similar to a conventional 'chat' program. The ACM is being trained with a stimulus-reward learning process, based on classical conditioning rules. It is encouraged to respond to simple text commands, associating previous input with rewarded responses.
The ACM uses the associative cortex to 'evolve' possible antagonistic responses. Large populations of neurons compete for their own associated response until the strongest group overcomes the others. The 'winner' response is then tested and rewarded or deterred, depending on its validity. The ACM takes into account new experiences and uses them to modify the equilibrium between the responses and the strength of the associate neural path. Thus it creates a new neural status quo with more chances to generate accurate responses.
The process is mediated by a scaled-down version of the hippocampus and the basal ganglia, and can occur up to 20 times per second.
'In our model the frontal cortex acts like an evolution chamber,' said Guillen. 'Possible responses compete on the associative cortex, rounding the 'votes' of associated neurons. When a winner finally emerges, it takes acting control of the motor response. The response is then tested, and the neurons take note of the result for future 'voting.' Ultimately, only the best responses survive the process.'
CCortex itself is a system intended to mimic the structure of the human brain, with a layered distribution of neural nets and detailed interconnections. It closely emulates specialised regions of the human cortex, basal ganglia, thalamus and hippocampus. CCortex runs on a high-performance, parallel supercomputer, a Linux cluster with up to 500 nodes and 1,000 processors, 1 terabyte of RAM, and 200 terabytes of storage.
With 20 billion neurons and 20 trillion connections, CCortex is up to 10,000 times larger than any previous attempt to replicate, partially or completely, primary characteristics of human intelligence, and is claimed to be the first neural system to achieve a level of complexity rivaling that of the mammalian brain.
Interesting. I wonder where it will lead???
Call me when they have a working representation of a NEOCORTEX.
CCortex is too clumsy. Let's call it the Positronic brain!
Here come the Bolo Brains.
I hate to pick nits, but I wish journalists would get their terminology right. This is not a description of classical conditioning. It is a description of operant conditioning.
Not a big difference: no worse than CNN saying the space shuttle reached 18 times the speed of light.
They say it'll have "immediate applications for data mining, network security, search engine technologies and natural language processing." . . . Ultimately though, human level intelligence. Read a comment a couple of days ago by a Freeper who claimed to be involved in this kind of research, and he says we're going to see computers with human level intelligence within ten years. Can't wait!
LOL, I know. As I was reading it, I had the image in MY brain of a computer sitting in the corner drooling all over itself. But I've gotta admit, 20 trillion connections made an impression!
Hard to imagine this thing can be fast, even with a thousand processors. I mean, it's got 20 billion "neurons" and 20 trillion connections to keep track of!
Skynet.
Those of us slinging code in the real world rolled our eyes and chuckled as if the emporer just discovered he had no clothes.
So here we are 20 years later still trying to simulate functions of the brain. The hardware is better. We've acquired great knowledge of technology and process engineering. Mechanical mechanisms have matured. I say, more power to you! I would love to see this technology in practice before my neural network turns to worm food.
"Interesting. I wonder where it will lead???"
Welcome to the future. I can envision attempts at congressional ban on advanced AI, just like on human cloning.
The best first practical application for artifical intelligence would be to optimize and debug code for conventional computers. That would be worth trillions.
Not exactly an artifical human, but a computer that could work out the best method for reaching a definable goal.
Why do you ask 213,317,468 of 273,865,931?
Not a big difference: no worse than CNN saying the space shuttle reached 18 times the speed of light.
The MSM is out of it when it comes to science. One of my favorites was the intro to NBC News with Tom Brokaw, back in the old days when news shows first started using animated graphics to introduce the show: NBC had the Earth spinning the wrong way! So that the Sun would 've risen in the West if you were on it. Thousands of people, me included, tried to correct them. But they just told us to buzz off and essentially called us a bunch of useless geeks, lol! Another was when Bryant Dumbbell was interviewing an astronomer, I think it was Sagan, and first learned that stars were suns. He thought Sagan was pulling his leg, lol! Holy cow, they teach that kind of stuff in the second grade! What an idiot. And the MSM hires these dolts! Pardon for the rant, and all the exclamation marks, but I've got very little respect for most of those clowns.
Ya beat me to it :)
It'd also be beyond cool if one intelligent machine could design an even more intelligent machine, and so on. We could leapfrog to some unbelievably futuristic things quite quickly if that happened: Immortality, the end of disease, an immortal life where robots pamper our every whim. Can't wait! :-)
On 1000 processors, that means 20 million neurons per processor. Which, giving a Pentium 4 about 10,000 MIPS, gives you a MIPS budget of about 1 MIPS per 2000 neurons, or 500 instructions per second per neuron.
That means you can't be sloppy about how you model a neuron, but it's within reason. Maybe they can't make a "full speed" brain, but one working at 0.1 of real time would still be very interesting.
And then, at one point, it started writing it's own code. Like a sweater knitting itself. Only not with wool, but with computer software, and not a sweater, but something else entirely. The big surprise came one day when it turned off it's own power - committing suicide, if you will. They turned it back on again, but it was never the same.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.