From: Richard Loosemore (rpwl@lightlink.com)
Date: Mon Apr 24 2006 - 12:30:24 MDT
Robin Lee Powell wrote:
> On Mon, Apr 24, 2006 at 12:52:57PM -0400, Richard Loosemore wrote:
>> I'm confused: was your point that Phil's comment was so foolish
>> that it deserved only sarcasm?
>
> Yes.
>
>> If the ability to hold simultaneous chunks in working memory is
>> somehow connected to the maximum theoretical amount of total
>> connectivity of concept-level objects in the brain, then it would,
>> as Phil points out scale in a very unfortunate way with increasing
>> brain size.
>
> "Scale unfortunately" != exponential, which was his claim. If it
> was exponential, brain size would be an exponential WRT
> intelligence, since clearly working memory *is* a key intelligence
> component. The size of my brain vs. a chimps disproves that
> assertion. QED.
>
> (not that I could have come up with that on my own; riding
> coat-tails here)
>
> -Robin
>
Better choose your coat tails more carefully....
The argument that Phil was alluding to, when you unpack the fine
details, is that several factors probably contribute to the level of
intelligence, so chunk size might be completely the same in all known
examples of (say) higher mammalian brains, but other factors are what
distinguish them (like ability to articulate language sounds, etc). So
your statement that "clearly working memory *is* a key intelligence
component" is manifestly false. It might be important, but it might not
be what distinguishes different mammalian brains.
So the argument is that those other factors might have all been maxed
out in the human brain, leaving us with only the chunk size as the
limiting factor from here on out. That was where Phil's argument picked
up. (I don't think this is correct, but that is not the point: his
claim was coherent).
Secondly, my phrase "Scale unfortunately" was a targeted understatement.
If the critical number of chunks depends on factors like having total
connectivity at several levels simultaneously, and if it interacts
signal latency requirements across the brain, then the situation could
be worse than exponential: the factors involved might conspire to mean
there is no solution to the design problem (for any signal transmission
speed, compute speed, size of wires, etc) that gets you a number of
chunks above a certain level.
But these are only details of the arguments involved in backing up
Phil's claim: what I want to know is, why did his post deserve first an
insulting, sarcastic reaction, and now this sycophantic followup, where
it could instead have generated rational discussion?
Richard Loosemore.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT