Hope wrote in message ...
>This debate (from what I have read of it, coming to it a little late)
>seems to have become rather aggressive. Let me try to avoid going down
>that line by suggesting that a lot of the diagreement here turns on
>misunderstandings and ambiguity on the nature of 'training'.
I would say that the trouble arises from the connotations of "training".
"Training" implies a trainer. It is possible to say that the universe trains
the brain. But having said that, what have we said. That the universe is
God? Or that God uses the universe to train the brain? It would be better
not to use the word at all when discussing the brain. Instead say that the
interaction between the universe and the brain leads to strengthened
synapses, forget "training".
>It is
>just the human ability to cope with these sorts of tasks which causes
>debate as to the power of neural nets as a model for the brain.
The neural net under consideration is composed of sensory neurons,
interneurons, and motor neurons. It exists in animals as a construction of
the DNA and is altered by experience. This neural net is not a model of a
brain, part of it IS a brain.
Possibly you are thinking of an artificial neural net.
>This is certainly a property of the
>human brain (I can call to mind a picture of a piece of paper with the
>necessary symbols written on it, and do the written sums in my head
>with these imagined symbols). The issue is an empirical one - can
>neural networks do it?
Since you testify that your neural net is able to do it, there seems to be
no empirical question left for consideration.
Ray
Those interested in how the brain works might look at
www.wsg.net/~rscanlon/brain.html
