In a message dated 12/11/99 10:24:30 AM Pacific Standard Time,
tripathi@statistik.uni-dortmund.de writes:
>Here, I would like to point out Regarding the Brain as Biological Model,
>I read in IEEE Computer (October, 1996, Daniel S. Goldin and Samuel L.
>Venneri, Ahmed K. Noor) --"...The brain is a million times faster than
>the fastest neural network we could dream up. A single brain cell can
>carry 10 times the information of today's most powerful computer --and
>it's more energy efficient..."
Not sure how the authors of that qoute derived the conclusion of the human
brain being "faster" than computers, but I would love to see a human do
differential calculus in milliseconds, or compute a principal components
analysis of millions of numbers in seconds.
This is all just for clarity...not for argument. The actual data points
directly in the opposite direction of that qoute. The brain is much slower
than computers. Computers just need to be programmed, the brain can do its
own learning which is sometimes a major problem.
The single neuron idea has long since been dismissed in cognitive science.
No one neuron contains any learning. It is usually vast and complex
systems of neurons that activate even the simplist of data. This issue is
often referred to as the "non-locality of information" argument.
Neuroscientists sought for years to find if memories or learnings were
located in specific parts of the brains. Meaning small localized areas of
sets of neurons.
I guess I get a tad passionate when I see "theories" that are antithetical
to large bodies of data.
Glen
[Host's Note: Doesn't the brain take advantage of much greater parallelism
that computers? That is, computers are faster at calculations, brain is
better at learning? ..Rick]
--Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>