Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!gatech!purdue!tippy!sawmill!mdbs!doug
From: doug@mdbs.uucp (Doug Waterfield)
Newsgroups: comp.ai
Subject: Re: Cross-talk, memory limitations
Message-ID: <1991Jan24.215249.6737@mdbs.uucp>
Date: 24 Jan 91 21:52:49 GMT
References: <9093@uwm.edu> <1991Jan23.050328.8357@mp.cs.niu.edu>
Reply-To: doug@mdbs.UUCP (Doug Waterfield)
Distribution: usa
Organization: mdbs, Inc.
Lines: 64

In article <8357@mp.cs.niu.edu> rickert@mp.cs.niu.edu (Neil Rickert) writes:
>In article <9093@uwm.edu> markh@csd4.csd.uwm.edu (Mark William Hopkins) writes:
>>
>>   What would happen if the brain began to run out of memory?  I tried
>>imagining that a while ago.  Basically what I saw was this.
>>
> We would grow old and die.

Why?

>
>>   We use a sparse coding to represent concepts.  If that coding becomes too
>>  .
>>  .
> You are making many assumptions about how information is represented.  I
>doubt that there is any justification for them.
>

Why?

>>   It happened to me a couple times recently, though not necessarily
>>  .
>>  .
>>  .
>>similar future events to happen.  The word here is "precedent", but I kept
>>coming up with "prerequisite".
>
> This type of behavior is often seen.  It probably occurs even more often in
>young people with presumably plenty of spare memory.  It is completely
>consistent with my (unpublished) model of the mind, but apparently
>inconsistent with yours
>
And again, why?  You refute Hopkins' arguments repeatedly, and the only proof
you offer is a model which you *don't* explain!  How are we to accept your
statments without even an attempt at proof?

>>   What struck me the second time this happened was the similarity of the
>>two words (precocious and precedent) which I had momentary difficulty
>>recalling...
>
> That similarity fits well with my model.
>

How convenient--the similarity fits the model which you do not describe.

>>   What was interesting about both situations was that they happened long
>
> Actually you mis-analyzed the situation because you based your analysis on
>an incorrect model.  But instead of recognizing it as a failure of your
>model you have added additional complexity such as assumptions about
>memory overflow.
>

So, when may we be enlightened by your "correct" model?

==============================================================================
Doug Waterfield		| Of course, the opinions expressed above are mine 
doug@mdbs.UUCP		| and not those of my employer.
mdbs, Inc.		|
------------------------------------------------------------------------------
"How many times do your hear it? It goes on all day long.
 Everyone knows everything, no one's ever wrong."
				-Rush, "Show Don't Tell"
==============================================================================