RE: Neuron Computational Requirements?

From: Billy Brown (
Date: Wed Apr 19 2000 - 21:50:33 MDT

Paul Hughes wrote:
> > However, I think you are wrong to dismiss indirect measurements based on
> > comparisons between the human retina and computer vision systems.
> I'm not wrong just on the simple fact that psychoactive drug induced
states of
> consciousness are directly correlated with neurotransmitter activity
within the
> neuron itself to say nothing of the larger neural network pattern. That
> is proof that there is *extra* computation taking place within the neuron.

It proves nothing even remotely relevant to the argument.

If we know that some particular set of neurons performs a certain function,
and we know we can duplicate said function with X MFLOPS, we have an
accurate estimate of the computational demands of replacing said neurons.

It doesn't matter whether the neurons do computations using chemistry,
quantum mechanics, or black magic. It doesn't matter how complex their
internal states are. It doesn't even matter if we have any clue how they
actually work. The fact that we can replicate their effects on the external
world with X MFLOPS is the only thing that matters.

If you want to argue against estimates based on this method, you need to
show that the sample tissue does less computation than an equivalent volume
of neurons from other regions of the brain, or that it does computations
that are somehow easier to duplicate on a computer than those of the rest of
the brain, or that the sample is somehow atypical in some other respect.

Going on about how complex the individual neurons would be to simulate in
detail isn't relevant. Once you've got an empirical measurement of the
amount of computation that they do, the question of how the work actually
gets done becomes academic.

Billy Brown

This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:09:38 MDT