Jill Bolte is a Neurophysiologist and she describes the human mind as a Bicameral Processor where the right hemisphere operates as a serial processor and the left hemisphere operates as a parallel one. In her book and lecture “My Stroke of insight” she articulates the experience of how a massive hemorrhage in her right hemisphere allowed her to experience what Buddhists refer to as enlightenment.(Bolte)i
The nature of consciousness is not completely understood; however the function of how the mind creates connections and thus “Programs” itself is well understood; Hillary Putnam in 1961 and Jerry Fodor’s in the 1960’s and 1970’s work on the computational theory of the mind articulates the brain as an information processing system. (Chalmers)ii
According to a number of sources the Human Brain weighs around 1.35KG and contains around 100 billion (100,000,000,000,000); Neurons (Elert)iii. Each neuron only has 1 axon but contains many dendrites connecting to the axons of other neurons; as stated within his site; the primary layout of the neural structure of nerve tissue in mammals is determined by genetics; however the nature by which neural tissue reorganizes itself is very different; given these basic constraints we may calculate the processing power of the human mind. However this is not without difficulty as we must simulate every molecule. (Debono)iv
IBM currently has a supercomputer to rival the power of the human mind; ASCI Purple and Blue-Gene/L meet the theoretical computational power of the human mind. (Henderson)v; however since Blue-Gene/L has never asked where it has come from we may assume that it is not self aware.
The two main issues with computers and mammalian physiology are that a computer chip is a finite state device; the human mind is not. The human mind with its 100 billion neurons has an infinite number of potential interconnections that are plastic in nature; it also contains an undisclosed amount of neruo-chemcials; each having it’s own nature of action and reaction.
“Neuroplasticity” is the brains ability to re-wire it’s connections according to function, it was discovered by Jerzy Konorski in 1948. (Konorski)vi The best example of neuroplasticity is the body’s ability to re-learn the use of a given limb or part after a stroke or injury to a specific area of the mind. Another facet of neuroplacticity is cognition.
The “Synaptic Self” outlines how our minds grow into the beings that we become as we age; this book on how the mind functions and grows states that the nature of organization of our neurons defines how we as beings exist and become self aware. (LeDoux)vii
Thinking Machines was a company that developed a system called the “Simputer” based on the doctoral work of W. Daniel Hillis and Sheryal Handeler during the 1980’s (it was founded in 1982). Thinking machines designed super computers entitled “Connection Machines” however the company was only briefly profitable before becoming bankrupt in 1994. The connection machines functioned according to models of the human mind and were excellent at various problems however due to a lack of funding a “Thinking machine” never came to be. Thinking machines pioneered work with “Hypercube’s” and laid the foundation for modern connectionless cluster based compute systems such as Blue Gene/L. (Taubes)viii
However two major differences need to be articulated to the problem of creating a self aware system modeled upon the human mind; Our mind uses not only connections but also chemistry and it comes somewhat pre-wired.
Therefore; if our mind were a computer then the software is our genetic code acting as source files with our objective or compiled binaries being the linked neural nets with their features equating to our neuro-chemcial interfaces and our since of “Being” would be a constant re-compilation of all code simultaneously all the time, whilst maintaining a fully functioning application and interface.
This dichotomy of state is the human condition; and it is also the reason very few concepts of cognition may be implemented by any given computer; such as strong A.I.; weak A.I., Machine Vision, Auto-pilots, et cetera; ad nosium.
If we were to create a computational system that had the equivalent of 100 billion chips each connected to each other in a dynamic net with audio and visual input and output processing that learned via feed back towards an objective of functioning with it’s surroundings whilst developing an ontological reference of it’s experience then perhaps we would have a system akin to HAL in 2001’s a space odyssey; however the technical challenges to this problem are gargantuan in proportion; the first of which would be memory.
Our memory utilizes the semi-understood hippocampus to encode long term memory into molecules; although this is a brash generalization our mind is indifferent to memory and actual experience; how then would we enable such a model in a computational system? (Holladay)ix
Would the system become self aware and if it were would we be able to communicate with it? If communication was one of the systems objectives and we gave it constraints based upon a reward based iterative learning mechanism with adequate models of interface and pattern recognition then yes it would communicate with us; however we would have to create a system flexible enough to understand the difference between noise, sound, speech, music and learn to “learn languages” first.
The potential is there and if we were to create such a system we could communicate with it; however we usually focus our efforts on creating massive systems that communicate with their own internal data; as such if a program ever became self aware would it even be aware of us? The blue mind project aims to do just this; create a truly strong A.I. that is aware and capable of full communication with us.
i Bolte, Jill (TED, February 2008) My Stroke of Insight [Online] World Wide Web, Available From:
http://www.ted.com/index.php/talks/jill_bolte_taylor_s_powerful_stroke_of_insight.html
(Accessed on November 17th 2008)
ii Chalmers, David J. (University of Arizona, n.d.) A computational foundation for the study of Cognition [Online] World Wide Web, Available From:
http://consc.net/papers/computation.html
(Accessed on November 17th 2008)
iii Elert, Glenn (Glenn Elert, n.d.) Number of Neurons in a Human Brain [Online] World Wide Web, Available From:
http://hypertextbook.com/facts/2002/AniciaNdabahaliye2.shtml
(Accessed on November 17th 2008)
iv Debono, Christian (Blue Brain Project, 2008) Blue Brain Project FAQ[Online] World Wide Web, Available From:
http://bluebrain.epfl.ch/page18924.html
(Accessed on: November 17th 2008)
v Henderson, Mark (Times Online, November 19th 2002) IBM Starts work on a computer to rival the human Brain [Online] World Wide Web, Available from:
http://www.timesonline.co.uk/tol/news/world/article833106.ece
(Accessed on: November 17th 2008)
vi Konorski, Jerzy (Journal of the Marine Biological Association of the United Kingdom, March 1949) Conditioned Reflexes and Neuron Organization[Online] World Wide Web, Available From:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1643801
(Accessed on: November 17th 2008)
vii LeDoux, Jean (n.p., 2002 ) The Synaptic Self
viii Taubes, Gary (Inc.com, 1995) The rise and fall of thinking machines [Online] World Wide Web; Available from:
http://www.inc.com/magazine/19950915/2622.html
(Accessed on: November 17th 2008)
ix Holladay, April (USA Today, March 12th 2007) How does human memory work? [Online] Available From:
http://www.usatoday.com/tech/columnist/aprilholladay/2007-03-12-memory-first_N.htm
(Accessed on: November 17th 2008)
No comments:
Post a Comment