Remi COTE 290196.1637
To Bill P.
Turing test are good, but they can be misleading when dealing with complex
system. If you can obtain the same result with and without integrator, there
is a validity problem. But I can sense that you don't are not "against"
if-then. If there is a discrepency then do something to reduce that
discrepancy. In some case system produce the same output, they just
modulate the frequency of these output, like hartbeat, if less oxygen, or the
bacteria (I forgot the name) that stop forwarding more often if less concen-
tration of a certain chemical. In some other case the output is modulated
by the magnitude of the error(with or without the use of integrator). I don't
see why the coexistance of the two architecture is to be rejected.
You say (960128.1800): "if temperature is less than 20, you must have one
signal representing temperature and another one representing 20. then you
must substract the first signal from the second one and see if what is left
over is greater or lenn than zero." not one signal representing temperature
but several, from several different source, that "some how" add themselve.
And if this addition is more than the signal ref sigals coming from different
source, then there is a firing of an output. Maybe one standard unit of
impulse like I naively suggested, or an error modulated output. So far
I uderstand. You add:"In a continuum system, however, a single neuron can
output a signal proportional (in frequency) to temperature error with an
effective resolution of 300 diff temperature." Are you saying that the
same neuron give out (output) a series of miniblasting that are viewed as
frequency? That is what you mean by current in neuron? If it is so, I want
to say that there is no need to argue. I agree with you on everything, and
I just hope I can find another question soon, because you are a great teacher.
And you are right I am lucky having an oracle that allow me contact of the
third type. (no second type)
REMI