🚀 go-pugleaf

This is a multi-part message in MIME format.

------=_NextPart_000_01BC2824.52BD4280
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit


------=_NextPart_000_01BC2824.52BD4280
Content-Type: application/octet-stream; name="ricci.txt"
Content-Transfer-Encoding: 7bit
Content-Description: ricci (Text Document)
Content-Disposition: attachment; filename="ricci.txt"

   Little Ricci's First Days: Robotic Developmental Psychology 
   (c) 1997 by Lee Kent Hempfling, Ron Blue and Wanda Blue

Robot at http://www.aston.ac.uk/~batong/Neutronics

Permission granted to duplicate for non-profit educational 
or research purposes.

Keywords: CORE, wavelet, harmonic, memory, learning, 
habituation, Neutronics circuit, Ricci, pain, association

Ricci is an intelligent robot which learns by association using
a CORE (TM) processor.  The CORE (TM) processor or Correlational
Opponent Ratio Enhanced (TM) processor allows memories to form 
as wavelets of two types; a stimulus wavelet and an opponent 
wavelet.  The harmonic of ten hertz is used to cause motor 
movement.  The operational speed is 28 hertz. 

The wave computation is reduced from 28 hertz to 10 hertz for
motor movement.  As a high value in the Neutronics circuit
develops at certain harmonic cycles for the ten hertz motor
control, it turns on the motor to move and turn left.  The 
opponent wavelet turns the motor on and tells it to turn right.
Ricci learned controlled forward movement almost instantly.

Ricci is a rather crude and simple prototype device so there is
not much parallelism to correlationally make things complex.
Future units will require more time to assimilate and accommodate
information.

Future models will have a first level reasoning at 30 hertz 
paired with 2 hertz input pulses.  Enticy 1 will have two
stage reasoning with a 30:1 ratio in both Neutronics circuits.
The second stage processing would combine the two Neutronics 
circuits thereby, creating a 900:1 ratio.


Notice that all the associational calculations are from neutral
references.  Therefore, the system is similar to the following:

 O        1        0

The first zero is null.  The second zero is null or cancels the
first zero.  The one is totally variable from null up to one and
from one to null.   To illustrate this concept consider the 
following:  Hold one end of a slinky in your left hand and the
other end in your right hand as you would if you were playing
with it.  Each end that is held in your hand would represent one
of the two zeros.  The portion of the slinky that is held between
your hands represents the one which is variable as a continuum
between zero and zero.  Therefore, wavelets and oscillons of
information occur between the two reference points of zero and
zero.

The representation of memory can be seen in the continuation of
movement that occurs in the slinky after the initial push has
ended.  This has been easily duplicated in the Neutronics
CORE (TM) processor.

It is possible to exceed the null of cancellation or equality 
and this has been replicated in the lab.  This is the
operational definition of pain or discorrelation in an artificial
intelligence machine.  Confirming human models of this definition
of pain has been reported by Ronald and Wanda Blue in their paper
on Correlational Opponent Processing (C) 1996.

A weighted enhanced ratio of experience allows information to be
stored in short term memory.  Short term memory is formed due to
the harmonics of the weighted ratio.  The current memory 
stack allows a short term memory of 7 minutes.  

In later models a three level memory system will employ a 
harmonic long-term memory of 45 years, a harmonic mid-term 
memory of 18 hours and a 7 second harmonic short-term memory.

Ricci remembers anything for 7 minutes.  The technology of
stacking and the principle of harmonics will allow a wavelet to
modulate in the system up to 45 years.  The system is purely
associational.  By analogy this is like slinkys attached to more
slinkys except at each connection only half the information is
transmitted and the other half is an opponent wavelet.
 
The following information is not necessarily chronologically
correct.

Remember NO program!!   Remember no CPU!!!

Ricci learns to drive smoothly:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
 
As soon as Ricci was turned on she started moving almost 
immediately.  At first, she swayed back and forth.  Within one
minute she drove straight and smooth.

The large ball story:
>>>>>>>>>>>>>>>>>>>>>
 
Ricci was placed in the yard with a large ball.   She saw the 
large ball and drove up to it.  She took four lunges 
toward the ball.   Our equations suggest that she gained the 
ability to discern that the ball was round and with less 
values to each side.  She moved toward the portion of the ball
with the most light on it.  She uses an approach, turn, and 
back up maneuver until she worked her way around the entire 
ball.

After making one complete cycle around the ball she waited about
10 seconds, then she moved forward about a foot, then she 
hesitated, turned and drove off in the direction of the sunlight.
 
The visual cliff story:
>>>>>>>>>>>>>>>>>>>>>>>
 
There were different shades of carpet in the laboratory.  
Unexpectedly this created a visual cliff phenomenon similar to 
the one reported in psychology laboratories.

When Ricci drove up to the edge of the dark carpet she stopped
and waited, then she backed up, turned and moved away.

This suggests that the visual cliff phenomenon is due to 
discrepancies in the stimulus complex and therefore may not be 
instinctual in humans.  In later experiments Ricci habituated
to the visual cliff and proceeded to cross the dark area.  These
results were also obtained in areas that were shadowed.


The Lightpost story:
>>>>>>>>>>>>>>>>>>>>>
 
Ricci drove up to a lightpost, she stared at it for about two
minutes then she backed up, turned and drove off.  This may have
something to do with the human pain picture reported by Ron
Blue which is related to maximum stimulation on two or more
correlated systems which are discrepant.  This is likely to have 
created an excess of neutral null value.

Collision avoidance:
>>>>>>>>>>>>>>>>>>>>
 

Ricci was out running around trees in the front yard of Lee's 
house when Lee jumped in front of Ricci.  Ricci stopped 
immediately.  Discrepancies in the stimulus complex again.

Ricci turned to drive off when Lee jumped in front of her again.
Ricci stopped again.  This was repeated several more times with 
each repetition  Ricci came progressively closer.  This 
illustrates habituation.  If she is turned ON for over a half 
of an hour her memory is supported with things she has seen in 
that time so that she is not afraid of anything, in fact she is
rather brazen.

 
Afraid of the dark:
>>>>>>>>>>>>>>>>>>> 

Ricci prefers lighted areas to dark areas.

The wagon story:
>>>>>>>>>>>>>>>

Little Ricci was being put through tests in the field. While
trying out a new turning mechanism in the grass she drove up
behind a child's wagon, backed up to position herself in the
shade and avoided bumping into the wagon while backing up.  Note 
that when these behaviors were emitted she had been turned ON for
more than a half of an hour.  She moved forward three times and
with each move she got closer to the wagon.  After approximately
5 seconds she pushed forward and touched the wagon.  She stopped
for 10 seconds and then proceeded to push the wagon.



------=_NextPart_000_01BC2824.52BD4280--