Physics
The expansion of cognitive matter occurs in three dimensions. The first two (implosion and explosion) involve creating new OSP entities. The third kind of expansion is parallel or "slightly imperfect" expansion, ie, trial mutations (usually minimal) in processing structure and/or event definition clusters (memories). (One may also wish to view the parallel expansion of an existing OSP as a new entity.)
Let a young entity explore his own ways and perspectives of thinking. In a mature entity, a multiple of sub-entities will observe and shape the three realms of a common task universe, each from a unique perspective, configuration, and perception of ETU m&m. Each module in the cognitive structure will locally discover and shape order toward entity objective, test unique processing sequences and formats, and try new points of view on its own three interactive realms of matter and motion, perhaps "independently" expanding from its local nucleic observation post.
This expansion may be viewed as a task oriented cognitive fission, perhaps randomly configured (in default), but adopted and maintained because of an increase in primary value associated with the observation and shaping of the designated portions of ETU.
Not only may an entity discover effective order more directly by shifting perspective and perception modes, but some order may be REVEALED EXCLUSIVELY from a given unique perspective/perception configuration (OSP). Indeed, I suspect a most direct approach to problem solving is to discover observer/shaper posts (within the three realms of ETU) revealing m&m patterns associated with value fluctuation in the target environment. (Notice that such a discovery is a recordable cognitive event that may repeated perfectly or approximately.)
Experimenting with additional observation posts on the existing matter/motion of the ETU, may commonly be the only avenue of change available to an entity facing an unresolved obstacle or perpetually elusive challenge.
The creator's final task is to establish conditions that allow the machine to evolve "naturally" toward its objectives, by allowing maximum freedom for: experimental change in LDM processing steps, selecting configuration and perspective expansion on ETU m&m. Design the CEM to continue to explore new observation posts within its three realms of entity task universe, until its assigned objective is reached.
More Observation Posts Offer More Intelligence
An expansion event can involve matter or motion in any or all of the three realms an the ETU. Indications of inherent environmental order vary from given observation posts within the three realms of ETU. The more observation posts that are considered, the more opportunity there is for discovering value related order. Notice that an adopted new decision or observer entity becomes part of the new entire existing structure (ETU) and may be configured in any future expansion events. - end reprint
Watching the Hand of a Thinking Machine
Consider the hand of a human being. This isolated mechanism senses and shapes the environment, just one small input/output peripheral agent of the entire somewhat intelligent human machine.
If you two-dimensionally plot the relative points of the knuckles, joints, and fingertips during normal daily activity... patterns reflecting inherent environmental order and actions are revealed. (For instance playing a song on a piano, would produce point-patterns profiling some of the order within the musical composition and the instrument.)
Plot these points in three dimensions, and more complex and revealing patterns occur.
Consider plotting the movement of all points as observed from the relative perspectives of each point.
Attempt to imagine a near infinite number of theoretical observer-perspective points. (points other than knuckle/joint/tip perspectives)
Finally consider the entire human entity defined as an observable mechanism, including a processing model of the working brain and its memory structures.
Six Aspects of a New Observer/Shaper Post
1) LDM learning circuit.
(I recommend beginning with a SFRSM.)
2) Selected Configuration Of The ETU
The matter and motion chosen to be considered in the observation may be from any or all of the three realms of ETU.
3) Sensory Perceptions of the ETU
The type and number of perceptions being considered may vary, limited only by the sensor capabilities of the entity.
4) Shaping Ability on the ETU
The type and number of specific shaping actions and the configuration of the m&m being shaped (Limited only by the shaping capabilities of the entity).
5) Perspectives of the ETU
Multi dimensionally defined particles, clusters, and mechanisms of the matter and motion in the model of the ETU will reveal different order, from different sensory perspectives.
The same matter and motion (when modeled multi-dimensionally) may be perceived differently from unique sensory perspectives. A perspective will be defined (and labeled) for each given unique observation of m&m.
6) Objective(s)
This may be a local concern, having a given priority based in default on the digressive "distance" from the primary ETU nucleus of the cognitive mass.
A cognitive physics entity cannot occur without a reasonably well defined ETU and an objective. The environment and objective are supplied by the creator of the entity (This may be a human being, or an expansion manager program). An immature entity shapes itself as a product of its chronological interaction with its environment. Like the mind and the body, a cognitive physics entity and its environment are inseparable.
An Example of Useful (OSP) Expansion
Within the ETU of a Traditional Car Creature
Automobile Entity with a Brain
Assume a learning machine (a SFRS for example) has been programmed to be the brains of an automobile, ie, a car that learns from experience and shapes toward standard passenger vehicle values. The general objectives of this entity are the common attributes associated with people moving vehicles, including reliability, passenger safety, and longevity of the AUTOentity's own structure (life span of the machine's parts).
I contend that by providing such a learning machine with certain given or additional perception, value enhancing associations become detectable, and further that such useful expansion may occur in a number of perhaps obscure places within the three part ETU of the this "car creature".
One Additional Observer Post
Consider two automobiles, identical except that one has a fuel gauge, and the other has no fuel gauge. If both AUTOentities are given FRSM brains, the one with the fuel gauge will inevitably develop greater "intelligence" because of this additional sensory input (observation post) on its ETU.
Consider an AUTOentity with the ability to independently install (over the entire ETU) additional observation/shaping perspectives such as the fuel gauge.
Watching fuel levels is one obviously useful observation post considering the general objectives of an AUTOentity. But there are useful, but less obvious, value revealing configurations of the ETU to be discovered. Supplying a learning machine with a new OSP (within a given ETU) may trigger the generation of intelligent reactionary behavior, perhaps impossible to discover before the creation of said post.
Monitoring of tire air pressure for instance (by a competent learning machine) would inevitably lead to longer tire life and safer running conditions. Both correlations are independently discernible by a typical LDM through the association of tire pressure with wear rates and passenger safety records.
Consider how an SFRS AUTOentity monitoring block temperatures would independently learn to recommend shutting off the ignition before reaching heat values that cause permanent damage.
Consider an AUTOentity that constantly measures the distance between itself and anything in its path, computing braking solutions for extrapolated collisions.
These are examples of AUTOentity (LDMs) developing intelligence (value related reactionary behavior) because of the existence of a given observation post on the Entity Task Universe. The point is, unless such observation posts are established, certain opportunities to detect and affect entity value concerns may not be detectable. Random creation of OSP within the ETU is the default method of revealing such elusive value related order.
Notice that many of these revealing OSP configurations may be comprised of matter and motion occurring exclusively within the cognitive space of an entity. A computer hosted LDM focusing on its own ETU RAModel may be viewed as an AUTOentity with the ability to establish gauges and perspectives at will.
The expansion manager of a given entity chooses "point of view" configurations (either randomly or by design) from the action and material combinations available within the ETU. The possibilities for unique observation points are vast, even within a comparatively simple challenging environment. However, certain basic ingredients of useful OSP configurations will normally be isolated early in the entity development, leaving the variables of these general recipes as priority areas for random or methodical "experimutation"
---
Creating Revealing PERSPECTIVES - Watching a Waterhole
Consider a waterhole located in an environment containing a multiple of entities that require water. Further imagine that there is a sudden change in the environment; and "intelligence" is going to be necessary for survival.
The specific change in the environment is... The shores of the waterhole become saturated, and turn to quicksand. When a creature approaches the waterhole, it is immobilized and subsequently sinks.
Further stipulate that the creatures in this universe are very shy, and will only approach the waterhole in solitude. From their individual perspectives, opportunities seem somewhat limited for the creature population to acquire this critical knowledge (event records reflecting this environmental order).
But if we install an entity to observe this waterhole environment, from this unique perspective, useful patterned order suddenly becomes obvious; ie, a cause and effect event sequence is repeated beyond random expectation (the default conditions for pattern detection).
COGNITIVE SPACE (a dimension of defined thinking machinery) is filled with waterholes (decisions or interactive processing points) that constantly have creatures (definition clusters) coming through for a drink (being processed). Some of these definable perspectives will obviously be useful; others will be attempted, tested, and if effective will be adopted by a given entity. The number of observer perspectives available (within a given cognitive space) depend on the variable combinations of cognitive matter (contained by said cognitive space).
Many environmental solutions may be reduced to "a reaction to a known pattern" by establishing these effective observer perspectives within the machinery of a given cognitive space. Such perspectives are discovered primarily through trial and success; unique overviews are selected either randomly or by previously designed methods of selection. Patterns are theorized, as Inherent Environmental Order, when they appear with a frequency beyond that of random expectation.
Such overviews may be of the processing structure and/or the event records of both the cognitive and target universe. Useful cause and effect patterns (records) may be accrued and isolated at any level of any universe within the entity or its environment.
Value remains a measurement in the 0-universe. Growth may occur from any SSM existing on any level in cogspace.
An explosion expansion may be created to oversee the entire (or an isolated portion of the) mechanism of an SSM. It views this structure, an SSM in cognitive space, as a tangible (shapeable) isolated processing machine, with tangible (shapeable) ADAs.