NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/ai/philosop/7153

Text File | 1992-12-24 | 6.0 KB | 135 lines

Newsgroups: comp.ai.philosophy
Path: sparky!uunet!spool.mu.edu!agate!iat.holonet.net!ken
From: [e-mail] (Ken Easlon)
Subject: Re: Searle's World and Computers
Message-ID: <[e-mail]>
Organization: HoloNet National Internet Access BBS: 510-704-1058/modem
References: <[e-mail]>
Distribution: na
Date: Thu, 24 Dec 1992 19:26:27 GMT
Lines: 123


Author's apology.  This is a corrected version of an earlier post which has
been canceled.

In article <[e-mail]> ,
[e-mail] (Christopher Ian Connolly) writes:

>In article <[e-mail]>, [e-mail] (Ken Easlon)
>writes:

|> For example, they don't teach you in school that the brain is the
|> dumbest organ in the body.

>I don't think they're likely to start, either.

|> Almost all of the computing power in the human body is at the intra-
|> cellular lever.  A very small percentage is at the macro-cellular lever,
|> and some of that is carried out by neuron firings.

>You're pulling our levers, right?

Well, maybe just a little.  But it worked, right?

>What about gap junctions in liver, the immune system, hormones, etc...??

If we define computing power in normal terms, say megaflops, I think it
would be possible to calculate a rough number for various biological
systems.

It seems to me that you can get a rough idea of the intra-cellular
megaflop number by looking at the bandwidth (as in bits per second) of the
communication channel connecting the nucleus to the ribosomes.

This calculation may have been done already, but if not, a number could
probably be derived by looking at the average daily nutritional requirement
for protein, figuring out the number of amino acid units used, and factor
in amino acid recycling.  Once we know the total number of amino acids
assembled into protein every day, multiply by three (nucleotides per amino
acid), multiply by two (bits per nucleotide) and divide by the number of
seconds in a day and the number of cells in the body.  I'll leave the
actual calculating for the more ambitious among us.

Once we know the bits per second being transmitted by a typical nuclei,
then there's the question of how many flops per kilohertz, an interesting
problem but probably not intractable.

Regarding inter-cellular communication, each polypeptide hormone
was produced by a string of DNA instructions, and thus some inter-cellular
communication is certainly a subset of intra-cellular communication.

I'm not sure about the ratio of enzyme production (intra-cellular
communication) to the rate of production of simpler hormones, or how much
protein production is required at the reception end,  but I suspect a case
could be made for all inter-cellular communication being a subset of intra-
cellular communication.

Also, I maintain, that macro-cellular computing rates could be calculated
from hormone (& neurotransmitter) transmission rates, and that
macro-cellular computing power is a subset of the body's total computing
power.

And finally, it should be obvious to the casual observer (I've been trying
to work that phrase into the conversation for some time) that inter-neural
communication is a subset of the body's inter-cellular communication.

I readily concede that converting transmission bandwidth to computing power
is open to a wide degree of latitude, and converting computing power to
intelligence is even more controversial.  Nonetheless, I think a case can
be made for the argument that neural firings represent a very small part of
the body's total computing power.

|> As we get older, brain mass stays relatively constant, but we keep
|> loosing neurons and DNA.

>Some neuroscientists  (e.g., Calvin and Edelman) argue that this cell loss
>actually represents an *accumulation* of information...the darwinian view
>of brain development.

Sounds to me like the old argument of the wisdom of age verses the speed of
youth.   It's probably a matter of scope, or the "effect domain" of the
type of intelligence we're talking about.  Young brains can probably come
up with more new ideas per second, but old brains may be more effective in
filtering out unlikely candidates.

If you're living in a rapidly changing world, rate of new ideas may prove
to be an important survival factor.  If you're living in a era of shrinking
resources, concept selection may prove more important as a survival factor.

In real life they're both important, you need a steady supply of new ideas,
and the capacity to select the best.

|> Surviving neurons just grow bigger.  This produces severe
|> bandwidth limitation problems in the cellular machinery.  One nucleus
|> with a limited capacity for transmitting instructions is asked to
|> service a larger and larger cell.

>You're claiming that the increase in cell size limits the "bandwidth", and
>then somehow you link that with the nucleus.  It seems to me that the
>nucleus-ribosome "bandwidth" isn't affected at all here.

I agree the nucleus to ribosome bandwidth isn't altered, but biological
communication doesn't stop at the ribosome, it simply get's converted from
mRNA to protein.  The next step is to get the proteins out to their work
stations.  If you have a fixed rate of protein production and a growing
number of sites (as in synapses etc) that use the protein, I maintain that
the bandwidth per site is being decreased.

>What *is* affected is the transport time for various important substances
>such as ion channels, ion pumps, actin and neurotransmitters.  Even so,
>this appears not to matter in normal brains.

Increased transport time (your concession) and decreased protein supply per
synapse (my argument) add up to slower response time.  Are you claiming
that neurite growth rate and synapse formation don't slow down with age.
I'm not saying that's abnormal, just slower (narrower bandwidth).

--
Ken Easlon                | "...somebody spoke and I went into a dream..."
[e-mail]           |    -Paul McCartney
Pleasantly Unaffiliated   |

-- Last response --