Message-ID: <034421Z14071995@anon.penet.fi>
Path: cantaloupe.srv.cs.cmu.edu!europa.chnt.gtegsc.com!howland.reston.ans.net!gatech!EU.net!news.eunet.fi!anon.penet.fi
Newsgroups: comp.ai.shells
From: an306358@anon.penet.fi
X-Anonymously-To: comp.ai.shells
Organization: Anonymous forwarding service
Reply-To: an306358@anon.penet.fi
Date: Fri, 14 Jul 1995 03:41:56 UTC
Subject: Overview of Cog Physics p8:10
Lines: 57

Re: Overview of Cognitive Physics (part 8 of 10)
(5000 word introduction copyright 1995) by D. Albert Harrell
   DAHarrell@aol.com

VII. EVOLUTION THROUGH PARALLEL MUTATED EXPANSION   
Including The Hyper-Evolution Of Contained Reproductive Entities  In
Competition For The Same Elements. 
   
 A. Perfect Parallel Clones Perfect cloning of an SSC unit or network would
be used in cases where the decision or  problem universe is well known and
defined, along with a know historically effective task (solution) network.  
  
 B. Imperfect (mutated) Parallel Clones.  Any cloning other than the above
perfect type. 
   
The genetic mixing that occurs when an egg is fertilized, is the combining of
two different entities, to produce an unknown unique third entity which will
be tested by reality.  This biological method of natural creation and
selection contains many useful parallel concepts for producing unique SSM
task networks.   
   
A specific task oriented SSM module appears to have the essential equivalent
of "genetic character."  This "reproductive information package" correlates
to the structure of the SSM network, and the individual policies (and Simile
Compare Decision SCCatalysts) for handling data of the decision points within
the given isolated SSM task structure.  This genetic coding (of the SSMs) may
be used in the "crossbreeding" and "genetic splicing" of two or more modules,
to produce a unique "offspring."  
   
The parent modules of these "hybrids could be chosen by human, SSM, random or
"natural" (effectivity) selection.  SSM task networks, with their ADAs and
SCCs, could be subjectively crossbred like dogs, to handle a wider spectrum
of task types.  
   
An "on line" module within a cyberspace model environment has all the
critical elements that are necessary for replication with evolution  (i.e.,
unique "genetically coded entities,"  interactively participating in the
shaping of their own environment, while reproducing according to the
creator/installer assigned "laws of natural selection").  
  
When confronted with an unfamiliar task,  which is unlike any in experience;
two or more SSM modules (perhaps deeply imploded task networks) composed of
task elements similar to those of the problem, could be selected and
crossbred.  These offspring might be tested at high speed on RAM dwelling
models, with qualifiers implemented on the SSM task reality.  
   
It would be optimum to have these entities follow an emulation of
"propagation by natural selection," (based on effectivity toward modular
objectives) with simultaneous development occurring at a multiple of points
in the task network.  
(cont. in part 9)

----------------------------------------------------------------------------
To find out more about the anon service, send mail to help@anon.penet.fi.
If you reply to this message, your message WILL be *automatically* anonymized
and you are allocated an anon id. Read the help file to prevent this.
Please report any problems, inappropriate use etc. to admin@anon.penet.fi.
