Copyright 2008 All rights reserved
ABSTRACT:
It has been suggested that understanding of biological learning can be used to develop AI systems. It has also been suggested that the brain is a special quantum computer. Previous efforts to develop a quantum computer has involved the use of simultaneous states or superpositions. gaussian matrix panassociative entangled quantum unification by bipolar transistors in opponent charge states "QUBOCS" can be use to form positive holes in a global/local "neutral" position to represent simultaneous superposition conditions. All the electrons/positive holes in a "neutral" charge coupled relative entanglement can be used to store self-organizing information in harmonic oppositional states of neutrality. Consciousness is an emergent property occurring in a global/local systemic memory resulting in a representation of a singularity or oscillon/perceptron.
DISCUSSION:
Conventional quantum mechanics suggests using the particle nature of electrons and their spin characteristics to represent bits of information. By using the spin of an electron and changing it from one spin direction to the opposite one can represent one bit of on or off. Superposition is on/off at the same time. It is well known that electrons simultaneously possess the qualities of wave and particle. Electrons in orbital positions around "neutral" atoms are quantitized in packets or jumps of energy. When electrons change their position they release or absorb quanta of electromagnetic energy or light as quantitized energy.
Quanta energy is released in bundles and these bundles can be collectively added to and subtracted from to provide a mean sum combination of quanta energy. This can result in a normal curve or probability curve. Many coins thrown concurrently will generate unique frequency patterns forming a normal curve or distribution. This relationship of one half is the key to understanding quantum mechanics and building a simple quantum computer.
Kane (1998) recognized that a quantum computer can only exist in a system that is strongly isolated from its environment and would dissipate no energy during the computational process. Under normal circumstances these requirements would be exceedingly difficult to accomplish. Kane provides evidence for the use of arrays of nuclear spins with locations in donor silicon. Kane proposed arguments that this could provide ways to perform computations independently and in parallel measurements on each spin in the array.
Consider the whole as one and the one is made of parts. Relative entanglement is another way of saying that the whole is greater than the sum of its parts. By starting with a large number of qubits, or all the possible configurations of spin states of all the electrons/positive holes in a redundant gaussian matrix/linear trap and Paul rf traps of NPN transistors, one can use the whole to form useful quantum calculations if the inputs are quantitized and the calculation is split into half. The deflection or pathways consist of two qualities – resembling positive and negative wavelet vectors. Information is always positive but interacts through interference. Information is never lost but at certain times and locations appears to be a supersymmetry zero, a high positive-vector fuzzy value approaching one, and a high negative-vector fuzzy value approaching one. In other words, supersymmetry zero is similar to information deflected into two directions - 1/2 positive-vector and 1/2 negative-vector quantitized information. The name QUBOCS was chosen to represent this reality of making calculations using interference wavelets resulting in zero or relative entangled quantum superpositions in positive holes .
The theoretical reasoning behind the QUBOCS approach was generated by the Correlational Holographic Opponent Processing Theory model of the brain and Hempfling's reduction to practice of that model to a CORE processor and the Neutronics Dynamic System.
Hempfling's (1994) efforts lead to the development of the Neutronics Dynamic System (Figure 4) by using a NPN bipolar transistor in a circuit with a photodiode and a simple 9V power supply. The NDS produces an output voltage that correlates with the intensity of light stimulating the photodiode without drawing any current from the battery. Light intensity is transformed to an electrical signal which is stored as quantitized information in a charge coupling configuration. While this bipolar process is currently expensive, efforts were made to use CMOS circuitry and a bipolar junction transistor of the PNP variety by Spencer (1997). The theoretical reasoning behind this approach has lead to some interesting applications in Bipolar Spectral Associative Memories .
By using NPN transistors as a charge coupling device, information exists in parallel in an relative entangled neutral state of positive-vector and negative-vectors in a neutral zone or QUBOCS. When measuring voltage outputs in the NPN version of the QUBOCS, it is necessary to state the measurement procedure. One method is the Negative Reference Method (NRM) and the other is the Positive Reference Method (PRM). The results observed are different because of the way information is reciprocally correlated. This becomes the operational definition of matrix algebra where A x B = C, B x A = D, and C is not equal to D. Matrix algebra is a well known property of quantum mechanics.
The method of measuring the voltage is determined by the terminal of the battery as a reference system. Negative represents the negative terminal and positive represents the positive terminal. The volt meter must be turned to the AC reading even though the power supply is a simple 9 volt battery. It is rather strange to observe this effect and the different readings due to the learned experiences of the machine. The Negative Reference method of measuring voltage is used for conventional circuits. The Positive Reference method is used for measuring QUBOCS voltages.
While voltage varies in the gaussian/linear trap circuit due to experiential quantumly encoding information in a Paul rf trap over a very slow clock speed, the actual information is globally stored as noise in a gaussian/linear trap and expressed in harmonic local zones. A harmonic local zone can have three values, a positive fuzzy value, a negative fuzzy value, and a superposition or neural interference value. The harmonic zones are like magnets in a field. If you move one it resets others. The memory is the current configuration of vectoral positions and relationships. These positions are neutral mirror reflections of each other in reciprocal positive/negative vectoral dualities. Information that is put into the system is quantitized so that the quantitized information interacts. This approach is a reasonable step forward from Robert Feynman’s and Albert Einstein’s ideas. Quantitized information is different from normal information. Notice that the system is set up in such a way that the circuits allow for classical confirmation of quantum fuzzy calculations.
The information exists until the next quantitized information is put into the circuit’s gaussian matrix net. Clock speed is set at a slow reference frequency or Paul trap frequency. Charge state memory of about 3.24 trillion qubits is updated at this slow clock speed and is gaussianly looped like a snake eating itself. The slow update allows the 3.24 trillion qubit to set itself relative to local and non-local positions. The slow update allows a long reference time in loop memory. The gaussian/linear time trap in the current model is self organizing, dynamic, and analog. A recent article suggest that quantum information can not be lost and will always influence the fuzzy outputs in the future. The memory is dynamic and is an updated memory. Classical rules suggest a particular memory will be lost if not re-experienced in the environment. Memory is a weighted history of the circuit past.
The memories formed are associational or correlated memories and they model how the brain forms memories according to the Correlational Holographic Opponent Processing Theory . The memories are formed and strengthened due to the interaction of in the QUBOCS circuit. Consider the interactions of electrons in a capacitor. In a simple circuit, two charged plates connected to a power source will generate a standing interaction of opposites. One side will be positive and the other negative. The only current flow is when the capacitor is charging to capacity and the voltage is removed. A capacitor then functions as a battery or a storage of informational energy. Such informational energy can be amplified and generate music in a radio or digital information for a computer.
Atoms are neutral because the electrons are balanced with the positive charge of the protons. A capacitor is neutral because the charge is balanced with a positive and a negative side. A transistor can be used as a special type of on/off capacitor. A large power input into a capacitor will cause it to explode. A large unbalanced charge into a transistor will cause it to explode. In a simple quantum computer the goal is to make calculations in neutral. The goal is to input quantitized information and allow the circuit to calculate its self organized neutrality over a period of time.
Transistors are quantum devices. A transistor has some important qualities that allow it to be used as a quantum computer. Consider, for example, the short distance between the positive charge side and the negative charge side; this allows the formation of very high capacitance. Also, consider the reverse flow of positive holes and the Hall effect as an oppositional balancing and calculating force. Also consider the high resistance to reverse current flow which is different from a traditional capacitor. Notice that it easily divides information into ½ and ½. Notice that it can also be viewed as a paradoxical 1/3, 1/3, and 1/3 system. There are two negative sides and one positive middle. Remember we are using the positive holes to function as a supersymmetry or superposition quantum storage of information. In other words the circuit is so arranged that the positive holes in superposition are redundant and can rebuild the information due to quantum relative entanglement.
It
is difficult to explain quantum information to people but I must try in
order for you to grasp the importance of these relationships. A
nickel when flipped will land on its head or tails. This
is ½ and ½ informational system. A
nickel can also land on its edge and remain there. One
observer on one side will see an head and
another observer on the other side will see a tail. One
might say one of the observers collapsed the quantum object or nickel into
a particular interpretation. It is
difficult to know which observer did this because time and causality is
reversible. Notice from this frame
of reference the nickel on its edge is in a superposition of tail/heads
leading to a strange relationships of 1/3, 1/3,
and 1/3. Quantum objects remember
all the details of chance and observations regardless of position in space
or time. All objects are in reality
like a quantum nickel.
Another example is a simple illusion of relationships. Consider the series (12) (13) (14) and (A) (13) (C).
The quantum object (13) can have many collapsed observations because it is entangling its information with many other quantum objects. The quantum object (13) remains true to itself but remembers its weighted history which may be observed as a gaussian matrix distribution with infinite tails. In the universe of numbers the quantum object (13) is perceived as thirteen. In the universe of letters the quantum object (13) is perceived as the letter B. In the universe of primitive people the quantum object (13) may be perceived as an animal. All frames of observation are correct relative to the observer. The quantum object “remembers” the choices in all the universes as a weighted history of probabilities. These weighted histories influence future observations. Quantum objects may also oscillate in time and space causing spooky actions at a distance.
Quantum information can be encoded on a transistor as a unique pattern of positions of holes and electrons in a transistor acting as a charge coupling device. That oppositional information is equal to about 3.24 trillion qubits in a standard NPN transistor that can be used to generate readable fuzzy values that can be used to control output in a self controlling and organizing system. Self control allows the possibility of consciousness. Thought begins in physical movement and self-control. Consciousness begins with a knowledge of our impact on the world or I am.
Information is oppositional or acting as an opponent to each other and is therefore in balance over the power spectrum of the machine. This describes precisely the state of affairs that the brain models according to the Correlational Holographic Opponent Processing model because of the nature of how stacking information in quantitized into oppositional steps. Information formed in the circuit consist of an activating wavelet or stimulus and an opponent wavelet formed from the weighted memory stored in the system. Notice there is a delay in the feedback of the negative wavelet or oscillation. Memory wavelets then filter or neutralize incoming information that is congruent. This allows information that is novel or new to proceed . The equation for one cycle of memory then resembles the following:
T= 3. 24 trillion qubits in one transistor
T=
1/2Tpositive + 1/2Tnegative
1/2Tpositive
= (3.24 trillion qubits - input stimulus)/2 + input stimulus
1/2Tnegative = (3.24 trillion qubits + input stimulus)/2 - input stimulus
Packets of information or harmonics of information and consciousness would resemble oscillons. (Umbanhowar's Home Page and movie of a oscillon is at http://super. phys. northwestern. edu/~pbu/) .
The efficiency of the QUBOCS processor in performing quantum computation is of value. Hempfling (1998) correctly recognized that current efforts by others have put more energy into the quantum computational process than has been received as information from the efforts. A ratio of energy in/energy out or 1:1 ratio would be ideal for a quantum computation system and this upper limit is reached by the QUBOCS processor.
In an effort to overcome limits in a binary environment, CPU speeds have been increased thereby providing an illusion and comfort zone that this is the correct way to create an artificial intelligent computer. Parallel computation also supports the illusion of progress, yet it is only simultaneous binary calculations. How effective can a system be if it only works on one calculation at a time? The goal in quantum calculations is the interaction of trillions of qubits at the same time. This goal has been accomplished in a QUBOCS processor. It is important to note that the quantum memory in a QUBOCS processor is significantly larger than its inputs. The inputs are processed as a whole. The inputs are only a small part of the whole. This seeded value then is ratio enhanced by using a Paul trap or a slow hertz clock rate in relation to stimulus input. The more congruent an input to memory, the more that memory will be supported in the entangled gaussian matrix distribution stack resulting in a larger amplitude of memory value and a system which learns by association. It will rebuild its memories any where in the circuit because it has a redundant quantum error correction protocol built in the circuit. All elements in the circuit would have to be seeded with errors greater than ½ of the information at the same time in order for an error to occur. Currently the only way this can be done is unplugging the machine or cutting a wire. An electromagnet pulse will not do it.
Why are we using these procedures in making a simple quantum computer? By understanding the basic processes the brain is using, we are able to duplicate those processes in an electronic circuit. To further understand, as simply as possible, these processes consider the following:
Ghahramani and Wolpert (1997) reports evidence that visual motor learning occurs through modular decomposition. Modular decomposition occurs when information is broken down into two or more variables. Proof of learning can be measured from the interaction of variables. Any learning task involves learning two or more variables at the same time. This is similar to Osgood's scale for measuring experiences. For example; the dichotomy scale of (good 1 2 3 4 5 6 7 bad). The quality of goodness is learned through experience and paradoxically the quality of badness is also learned through experience. A feeling or behavior can be measured by mixing modular components. Modular expert neuro specialization areas created by learning would send hierarchical gaussian matrix mixtures to create generalizations or multiple relationships. The central limit principle and the resulting gaussian matrix normal curve is a natural consequence of summing interacting relational information. "This relationship results from the assumption that each expert is responsible for an equal variance gaussian matrix region around its preferred starting location, which corresponds to its receptive field. " Expert neuro specialization areas using the good/bad dichotomy is illustrated by a receptive neurological field for good and a receptive neurological field for bad with both sending a signal of their respective weights for integration. gaussian mixtures allow calculations. All calculations or possible relationships are calculated through the interaction of the receptive field. The integration or conscious/unconscious singularity/duality CHOOSES the application for the current stimulus situation from wavelet interaction with the stimulus. There is no special or hypothetical area deciding which choice is made. It is determinism and probability mixed together. In other words, it is not necessary that you experience everything to know a particular case. It is global and local. Consider the case of modulating a fountain pen in front of your eyes. It appears to be made of rubber. Now put the modulating pen in front of your computer monitor and you will see that it appears to be made of particles or multiple pens. This illustrates a perceptual manifestation of modular interaction. The modulation model was a significantly better fit to describe the observed behavior than a linear model. A linear constraint model would predict a linear generalization pattern. This was not confirmed by the data. The visumotor system has limited generalization to novel events which suggests local receptive field structures. The experimental results "show that learning two new visumotor mappings, whether represented as vectors or postures, at two starting locations, leads to a smooth sigmoidal generalization at intermediate locations. " Meaning comes from experience and interaction with one's current situation. The strange thing is that you do not have to know to learn. The research may be interpreted to support Correlational Holographic Opponent-Processing for the following reasons: gaussian matrix receptive fields model is supported, modulation interaction that is sigmoidal supports wavelet interaction, simultaneous multiple learning supports global interaction and local interaction, and visumotor areas are functioning as activating and inhibitory centers. All these events are a normal consequence of the wavelet nature of neuro processing . The neuro structure is a global history of previous and current environmental stimulations. Behavior is never dependent upon a single neuron.
This process is simular to the formation of physical oscillons in a vibrating system with two frequencies ( Umbanhowar, Melo, and Swinney 1996 ). Oscillons modulate and exist due to the unseen wavelet interactions of the two frequencies and the history of the system. Oscillons are the observable memory in a vibrating system. That memory is made up of a positive particle phase oscillon and a negative particle phase oscillon. This particle oscillon can be thought of as a figure and the apparent noise oscillations around the oscillon as background. Notice memory consists then of figure and ground, local and global, long-term potentiation and long-term desensitization, short-term potentiation and short-term desensitization with all modulating in time. Memory then is dependent upon reference frequencies, stimulus overwrite on that frequency from a sensory field, correlational opponent filters, oscillating oscillons created by interaction wavelets by using neurotransmitters and evoked potentials. This models the quantum dilemma of particle and wave at the same time.
Vannucci and Corradi
(1997) at the University of Kent at Canterbury, UK have written an interesting
paper that relates to these issues. The paper concerns wavelet shrinkage
techniques through orthogonal and linear wavelet transformations, which allow
decomposition of noisy data into a set of wavelet coefficients so that noise can
be removed by shrinking the coefficients. The QUBOCS processor uses similar methods to form wavelets and oscillons and
to reduce noise by simply dividing the information into oppositional halves.
One half of a stimulus history interacts with current input data which
forms a new history. Stable oscillons and wavelets of memory form from
this interaction. The Bayesian model is a summation
statistical model with a mean of zero with gaussian matrix high and low bypass
filters for wavelet extraction. This also describes the QUBOCS processor.
The mother wavelet generated by this Bayesian model suggests why
a conscious robot, can have self control and self directed behavior. Mother
wavelets would represent, from a philosophical point of view, an idea or
correspondence to a schema in the environment. Visual symbolic representation
of this is suggested by eigenfunction
pictures . Covariance structures of
random wavelet coefficients allows learning and creativity to occur.
The reason being that any mother wavelet must have harmonic wavelets of
lower strength. This supports Ghahramani and Wolpert's (1997) conclusions
and observations of generalization and modulation in visumotor learning.
Additional research is suggested by Vannucci and Corradi's report of using
BayesShink on blocks, bumps, heavisine and Doppler signals seeded with gaussian
matrix white noise. The BayesShink is successful in recovering the data
with the exception of the Doppler signal. The Doppler signal is distorted
at the beginning of the signal. This wavelet interpretation of neuro processing
should demonstrate problems in a Doppler signal and allow a way to find
problems in the model.
REFERENCES
Blue, Ronald & Blue, Wanda. (November, 1998 ). Correlational Opponent Processing: A Unifying Principle. The Noetic Journal
Ghahramani, Z. & Wolpert ,D. (1997, March 27 ). Modular Decomposition inVisumotor Learning. Nature pg. 392-395.
Hempfling, Lee Kent. (1994, 1996). The Neutronics Dynamic System. Enticy Press.
Hempfling, Lee Kent. (1998). THE ROTATING TURTLE. Enticy Press.
Kane, B. E. (1998, May 14). A Silicon-Based Nuclear Spin Quantum Computer. Nature vol. 393 pg. 133 Spencer, Ronald G. (1997, November). Exploring the Use of PNP Bipolar and MOSFET Transistors in Implementing the Neutronics Dynamic System. Enticy Press.
Vannucci, M. and Corradi, F (1997, May). Some findings on the covariance structure of wavelet coefficients: Theory and models in a Bayesian perspective. unpublished report UKC/IMLS/97/05
Umbanhowar, Paul B. ; Melo, Francisco and Swinney, Harry L. (1996, August 29). Localized excitations in a vertically vibrated granular layer. Nature p793-796.
PROPRIETARY TECHNOLOGY
The material detailed in this paper is covered
by filing status with the
Figure
one (Hempfling, 1998)
"(Fig
3) provides a look at a simple connectivity of two independently variable
pathways, from the same power cell (same frequency) which are coherently
blended to form a single variable output. (This example is for demonstration
only and is not employed in the device design). Same frequency outputs
from A and B are joined in the same configuration to create output C, which
is coherent while being nearly environmentally impervious. The notation
is (((A-B)/2)+B)=C. (Hempfling, 1998)"
"(Fig 4) depicts the operational parameters
of the anyon system, termed the Neutronics Dynamic System.
figure
four (Hempfling, 1998)
