071b-Approaching a basic model [20120714]

the representational chains/branches are where purpose and "action" come from.   it is the representations which drive the branching.  
the stigmergic messaging serves the purpose of representation chaining.

thus we have two systems here:

there is a representation system, which can be thought of as the representation branching process, where representation pathways are being processed and active, these are the objects of awareness. 

the other system is the actor/structure system.  it is the stigmergic messaging and actor/automata system.  it instantiates the whole representational world.  The active parts of the representation system will have active parts in the stigmergic/actor/automata system processing those representations.  New creations in the representation system is managed by stigmergic/actor/automata processes.

in biology, this actor/automata/stigmergy system has it's base in chemical and possibly quantum effects.  The narratives of representational processing push down into actor/automata/stigmergic processing in 3 ways: 

1) energy  2) messaging  3) entities.  entities may be clusters of cells communicating with other cell clusters or organism to organism.  but most entity processing is really just message passing/processing.  

and all this processing is reduced to facts about space, form, and transformation.  the cells exist in a space, and communicate in a space.  the form of cells and chemicals determine how messages and energy are handled.  transformation is managed by stigmergic signaling (at best).  there is no higher order control of cell level energy and messaging.  it's all cellular level events.   But these cell level events are manifesting a simulation, a global representation that directs cell behavior (messaging, growth, addressing, etc).  

in the computer, the structures of stigmergy/automata/and actors and their addressed connections do not exist in a space, and have no energy requirements, many of the chemistry features of biology simply do not apply.   so how can a computer engage in that stigmergic cell process and produce a global simulation and make representations?  

in biology we have a representation space above or manifested by the stigmergic/cell space.  The stigmeric/cell/chemistry and rep/simulation space interact with each other.


because in a computer there is no actual stigmergic/cell SPACE to develop those processes in, what do we do? 

we create a representational space below the stigmergic/cell structures in a computer.  

The computational system is a representational system that manifests the cellular/stigmergic space.  the cellular/stigmergic system then interacts and manifests with the representation simulation.

we don't have to model the same cellular/stigmergic system as biology.  and in fact, the cellular/stigmergic system is probably going to overlap with the computational system in a variety of ways.  But here is the key fact.  as programmers and developers we do not try to model the representational simulation.  we do our programming to create a cellular/stigmeric system that CAN make a representational simulation.  

We can use evolutionary tools to develop such a system, but it seems like a useful and fun endeavor to produce the code and axiomatic system that produces a cellular/stigmeric system. 

something to remember about any cellular/stigmergic system is that it is self contained.  there is no (unless it is using energy as biology does) creation from nothing in the C/S system.  Representation goes all the way down.  The C/S system is doing representation too.... it's just that awareness is always on the level of representation above the stigmergy.  This is a feature of consciousness because the simulation is the being that is aware.  it is the repper.  

the branching exists in rep simulation, and the C/S system is what manifests the rep simulation branch and action structure.  

When we develop below the level of C/S system, we can make new signals that can be put into the C/S system and used.  each C/S system then develops it's own representational system of branches and simulation using that new signal... each object in the C/S system exists to help it do representations.  

the C/S system is not about solving computational problems, it's about solving representational problems. the level below the C/S system, is about solving computational problems to develop signals/cells/actors/automata.  These then naturally can solve representational problems. 


#20120715 basic model part 2  

The big problem of representation is this:  how does representational branching and chaining affect the cellular/stigmeric system?

the short answer is simple.  Representation is always happening.  That is the key thing.  There are always representations being made (the up arrow).  And these representations produce changes to the C/S system (the down arrow).  

the representations that get created MUST be nested in a representational structure... the web or branches of representations.  we can add arbitrary leaves and branch points into the representation simulation, but it's all the connections that dictate which branches, which paths get used.  it's the complexity of the representational branching (searching) requirements that is how the representational simulation tree is traversed. 

and to be clear, for processing, we do searching.  we do searching with all the variables (from point to point) in the representation simulation space.  and if there is no match, we start dropping variables until we have a match (we move up or down the tree for an alternate path).   we are looking for the right address to connect to in the C/S space.... and if there is no satisfying address, we drop requirements until one is found.  The C/S space is doing pattern matching, where the patterns get reduced if there are no matches.  

note that at the point we go back up the representational tree, that is the point, in the C/S system that we must start growing new addresses. 

there are other variables,  we can search longer, or we can add variables to restrain the results.  but it's a pattern matching/search problem in the representation simulation and C/S system does a corresponding address based signal processing. 

now, because representation is always going on, sometimes it's the new connection that gets used, that gets processed.  

and new representations are what in the C/S system?  they are new groups of signals (a signal pattern)   or they are new connections between cells.   or they are newly "charged" or  "accepting" connections between cells.  remember, the cells inhibit and excite each other.  this aspect of stigmergy plays out in how representation processing happens, and in representations "coming" into being.  

to be clear though, representations must be cellular connections, or cellular connection clusters.  until the C/S cells and their connections exist, the representations they "endow" or "instantiate" do not exist, or cannot be accessed.   it's the connection architecture of the C/S system that creates the representation architecture.... and vice versa.

and we are always making representations near the things we process.  some resource of signaling, of materials that make up cells, is put to making connections of things we "think about". the rep space makes new C/S connections ALL THE TIME.    how those survive or thrive is a different process.  

we make the reinforcement mechanisms too... and to be clear.  every thing is in the representation simulation space.  an inhibition to some behavior (like inhibitory response, such as inhibiting inhibition to do improv) is learned... it's a representational construct of rep trees, branches, paths that is part of awareness... and those things are C/S processes.  the path is making new C/S constructs from the RepSimulation space.


#20120718  basic model Part 3 - no halting - interplay between c/s and repsimsys:

I've been thinking for days, and dreaming too.  getting dream time where I get so close to grasping the idea of two systems interacting with each other. 

sometimes i see it as a web page, that is fed by two servers, and has some process that reads the webpage, as a browser, and then feeds it's answers back to those webservers.

how does the interaction work between the rep-simulation system and cellular stigmergic system?

that and I wish I had someone to talk to about my ideas and their implications.  ranging from to rationality and religion to being and body.

perhaps the one system tries to keep the other running.

they simulate each other... they try to keep the other from halting.

if one system halts what will it do? 

how do you un-halt?

is halting signal cessation on the CS side?   if signals cease, what happens? 

so many representations are stigmergic signal generators.  is that related to halting, to imminent halting?

There must be intermediate levels of interaction between the repsimsys and the cell/stigsys where a signal in the c/s system has a value in the RepSimsys.

ie  feelings, emotions, bodily sensations.    the organ/rep boundary.

--How does the c/s system go about making reps? 
--How does the repsim system go about maintaining/developing the c/s system?

-- The c/s system builds connections of neuron cells to make reps, and to make reps - signals, branches, processes, paths, loops, etc.    It makes reps ALL THE TIME. 

-- The repsim system runs the show.  it tells the c/s system what to do via actions, feelings. 

the repsimsys takes inputs from c/s and converts them into actions and feelings.  the repsimsys makes feelings, it makes actions. (these are reps obviously)

reps (brain structures+signals) make stories?  that cause actions and feelings - direct stigmergic signaling.

representations are VERY detailed.  rep process paths have all kinds of details that must be in plac to activate them.  to run the rep, to experience the rep to remember the rep.  the smallest variation (of signaling) can SHIFT an experience to not being the same.

it is a big sameness/difference game that is the basic search/like process going on. 

the "search" happens at the branch.  same as this address or same as other address.

a cell gets messages and then sends messages the stigmergic message summation (sic) sends a new stigmergic message on ether globally or locally (or both).   many clusters of cells/actions are message originators or need only a single particular message to start them up as message originators (the itch)

note that the immune system tries to keep the organism from halting from invaders.  it is a protective/signaling role.  the immune system weakness is it can turn on the host or platform of the immune system.    but what you see between the representation simulation system and the cell/stigmergy system is that they are VASTLY different, but they are structures of each other.  the nervous system structures ARE ideas.  the ideas are nervous system structures.    it's only understanding that it is MAKING representations in both systems that we have development.  or conscious development, learning, understanding, perspective, ability. 

when we look at our computational devices, we model and understand and program them as algebras.  but if we seen them only as machines, they are not doing any algebra....  a true AI must make representations, and the representations must change the AI signaling/structure system. 

the basic function for both the rep sim system is keep the signal/structure system (C/S system) from halting... because it can know what a halt is.  the immune system does not represent a halting state.  the rep sim sys represents halting states and tries to avoid them.

of course, halting in an organism means dying.  the central nervous system does input;idea;output functions.... expressly to NOT DIE. thus ideas are representations about state.... which looks like computation (and it may be) but the state is not about the representation system, but about the C/S system.   the representation system represents the whole organism, even if the representation system is instantiated by the central nervous system+ other stigmergic signals. 


#20120729  Basic model part 4  fuzziness

Apostolos Syropoulos is a genius.  hypercomputation  book is great.  fuzzy chemical abstract machines, is good too! 

how does the C/S system (an automata system) interact with the RepSimSys? 

the two systems have a fuzzy interaction.  essentially a representation in the Rep system is some degree of that representation.  and the stigmergic signal in the C/S system is some degree of a signal.  except that the those two fuzzy elements are fuzzy in the opposite system.

in the RSS, a rep is not fuzzy at all.  but in the C/S system it is the fuzzy value of some signal to cause some action or computation.  in the C/S system, some signal is a fuzzy value of some representation. 

the representations, on both sides, are the architecture of processing.  the stigmergic signal in the C/S system is some fuzzy degree of a rep in the RepSimSys.  And the RSS generates a fuzzy degree of some stigmergic signal. 

this gives us the interaction of "thinking" and awareness between both systems.  but remember.  architecture of the C/S system addresses ARE the representations. 

we can use this same fuzzy degree of connection to build and maintain the architecture.

for architectures to work, it must mean that the representations themselves work.  the signal processing of thinking through an architecture is about the representations producing C/S fuzzy degree signals to instantiate signal making by each cell.   


#20120803  Basic model - part5 re-explain 

Here are the postulates:

The rep simulation system is real.  

Each representation corresponds to a C/S structure  R ≈ C

A representation produces CS signals ≈ to the degree the representation encompasses experience - to maintain homeostasis of the organism.  Note:  THE STRUCTURE PRODUCES SIGNALS.  When the structure is activated, it will produce signals to other cells.  both via addressing/connections and in solution. 

Cells produce signals (to other cells) to maintain homeostatic function - of the representation (C/S) structure.  

the degree of signaling in connecting cells is representationally more like / less like.  the point of these signals in a structure is to "activate" it.   

to develop C/S structures we see the same kind of functionality, with different signals.    the cells are trying to create a certain kind of structure, stigmergically (bottom up)

The C/S cells signal cells to maintain homeostasis of all structures, to maintain the groups and structures of cells. 

C/S structures are cell groups.... tissues.  signals are representations of a group condition - quorum sensing.  the structures are interfaces of two simulations.  The C/S group simulation (the structure is the stigmergicly created form of the group) and the rep simulations that exists to maintain homeostatis for the organism.  

Cell networks look a lot like petrinets with tokens where the tokens have proportional values/degrees.  
previous next