In article DHtqv7.4wK@uns.bris.ac.uk,
eecjr@zeus.bris.ac.uk (CJ. Railton) wrote:
>
This is a disturbing thought which keeps cropping up, at least
to me. Mainly because of what seem to be the striking
similarities between the situation and function of a neuron in
a neural network, whether that be an Artificial Neural Network
or a human brain, and that of an individual in society. The only
fundamental difference seems to be that the neuron, especially
in an ANN, performs a simple deterministic function
Where did you get the idea that a biological neuron performa a simple
deterministic function ? Can you point to an instance in which the
output of a neuron is predictable for all possible inputs from neurons or
other input devices connected to it ? For most
possible inputs ? For even a limited range of inputs ?
whereas the
hyper-neuron, which is the individual in society, is complex and,
arguably, non-deterministic. If we consider a complex ANN whose
cells are made up of simpler ANNs and we add some random noise
to the system, then we have, in principle, something eerily close
to what we experience as individuals in society. For the purposes
of this argument, issues of consciousness and teleology which
although are connected, are not directly relevant.Drawing some analogies, we can say the following:
1: A neuron, in a brain or in an ANN will produce an
output which is the weighted sum of its inputs.
Biological neurons more closely resemble non-linear amplifiers or
attenuators, not simple adding machines. Adding machines do not
generate exponential progression to a state-changing event, which
is what happens in neurons.
>
An individual will largely absorb the cultural influences, the contemporary propaganda, philosophy etc. and express opinions which, despite how it might seem to the thinker, appears to the observer to be, sometimes very predictable, weighted sums of the input.
Siblings raised in the same household by the same parents in a stable
environment would be expected think like peas in a pod. The varieties of
outcomes among siblings suggests that your assertion overstates reality.
The extent of individual, truly original thought really seems to be the exception.2: Throughout the development of more and more complex
life-forms there has been a tendency for the control
functions to become localised within each organism,
culminating with the stage where almost all control
functions are carried out within a dedicated organ
such as the brain.
This may be true for control functions involving neurons,
but it is not true for the vast majority of biological control
functions. They occur in individual (non-neuronal) cells at the
level of molecule – molecule interactions. Many control signals
are disseminated via non-neural paths: hormones taveling in bodily
fluids; gases which permeate all compartments of an organism,
short-lived molecules which self-destruct rapidly, so must be
continuously produced to prevent a change of signal, etc.
>
More recently, in evolutionary timescales, there has been a tendency for control functions of life on Earth as a whole to be concentrated in ( usurped by? ) a single species, ie. humans.
Are you confusing control with destruction ? Granted, we plant
seeds so flowers, food and forests grow where we want them to grow,
but the seeds largely do their things while we watch and keep our
fingers crossed that things will go well. That is not control.
The only law our “control” seems to obey is the Law of Unanticipated
Consequences.
>
There are other analogies possible when the processes in the NN
and in society are looked at more closely, such as the way ideas
and political or philosophical movements “propagate” like waves
through a population, and the ways in which a NN responds to a
change in stimulus.Looking at it another way, suppose we could put ourselves in the
place of a neuron in an NN. What would we see? First we would be
linked through some communication mechanism to a finite number
of our neighbours.
Communication in a neural net probably is not limited to input from
neighbors. Distant signal senders can get their messages transmitted
essentially transparently through intervening sites by appropriate signal
conditioning, just as it is done on the Internet. In complex biological
settings, transmitting a chemical message in distinctively timed waves of
emission is one way to do it, analogus to digfitizing data being
transmitted electronically.
Overall, I think you are raising an interesting point; we have been linking
almost all humans together through modern communications systems. Our
thinking and behavior is being coordinated in both timing and content
world-wide. We may indeed be in the process of profoundly altering our
behavior as a lifeform. Maybe we are on the way to becoming more like colonial
instects – termites, bees, ants, etc.
A big difference in eastern and western cultures is that Western cultures
have never learned to deal with the absurdity of the limited lifespan of the
individual: Why do we live if we are going to die ? Darwin gave us the answer:
individuals do not survive; species survive. Our only biological raison d’etre
as individuals is to increase the probability of the survival
of our species. While people from eastern cultures also prefer to live, their
cultures seem to better understand that it is the culture and the species –
always the next generation – which really matter in the long run, not the
mortal individual. Each of us carries only a sampling of the gene pool of the
species, and cannot self-replicate. For those reasons alone we as individuals
are not representative of the surviving form, but we in the West manage to
overlook this fact in almost of our decision-making. As individuals we are not
properly wired for “controlling” all life on earth.
Thoughts couresy of Phil Filner