A problem for "Web as distributed programming paradigm"

Robert S. Thau rst@ai.mit.edu
Fri, 24 Aug 2001 13:45:32 -0400 (EDT)


Jeff Bone writes:
 > > To take one obvious example, the huge preponderance of the
 > > chemical work that goes on in the brain goes on inside the neurons
 > > themselves, while the information that passes between neurons is all
 > > done with just a few ions, Na+, K+, and Ca++.
 > 
 > Holy shit, Clay.  Biology is REST-based?  Is that analogy just completely
 > broken, or have you just made a completely profound observation?

I'm not sure I see the analogy, but in any case, the brain is more
complex --- information transmission between neurons relies on a vast
witches' brew of chemicals, including amino acids and modified amino
acids, small peptides, modified metabolites, and even nitric oxide,
which is released by neurons and propagates to their neighbors by
gaseous diffusion.  Some neurons also have direct electrical
connections to neighboring cells, called gap junctions.  

Also, you can't necessarily consider the output of a single neuron in
isolation --- the timing relative to other neurons may matter.  For
example, there's a rhythm in the "mental map" found in rat
hippocampus, in which (to oversimplify) the firing order of the cells
is correlated with the direction the rat is moving in[1].  (I'm not
pretending to know what this "means", BTW --- there's probably a Nobel
prize at stake for the first really persuasive explanation.  But it
is repeatable[2], and seems likely to have *some* functional
significance).

Lastly, it may be that cells other than neurons are involved; most
cells in the brain are things called glia; these were long assumed to
have no role in information processing, largely because they don't
form synapses with neurons.  But they do form gap junctions.

In short, the one thing we know is that this is a complicated, messy
system, and the organizing principles at the cellular level are rather
poorly understood at this point...

rst

[1] The simplest of the complications:  this assumes that the rat is
    moving through a familiar environment, and it also depends on the
    form of the environment to some extent.

[2] A similar pattern is observed even in sleeping rats, which was the
    research behind the flurry of "rat dream" stories in the popular
    press a few months ago.