[extropy-chat] Can a single brain cell recognize Bill Clinton?

[Adrian Tymes]

--- "Terry W. Colvin" <fortean1 at mindspring.com> wrote:
> (1) How can single, ostensibly rather simple, neurons process the
> flood 
> of bits
> arriving from subject's optical system?

As you noted, it might simply be the end cell for a bunch of other
processing.  It could, say, filter the values from other cells that
recognized eye spacing, other cells that recognized lip thickness, and
so on for other characteristics - many of which are used by the better
computer facial recognition algorithms today.

Unless all or most visual neurons link directly to these cells (which
would be easy to notice and remarkable in itself, given the unusually
high number of input synapses they would possess), something like this
would have to be the case.  True, one can imagine data from multiple
visual neurons being encoded into single-neuron inputs - but any method
that did this while preserving the information might not be
distinguishable from the encoding doing the preprocessing.  Every
neuron is a computational element of some sort, and there's a limit to
how far a single neuron can transmit its data.  Thus, information can
not get from one part of the brain to a relatively distant part without
potentially going through quite a lot of processing.

> (2) Can Darwinian evolution account for single-cell pattern
> recognition?  Of
> course it can; it *must*!

Kin recognition, mate recognition...there are quite a few social
functions that evolution would easily capitalize on, and homo sapiens
is definitely a social animal.

Which is not to say that Darwinian evolution accounts for all things,
at least not in the strongest positive sense.  It is arguable whether
evolution "accounts" for vestigial things like a human's appendix, even
if it can explain how such things came to be.