Newsgroups: alt.philosophy.objectivism,alt.sci.physics.new-theories,sci.physics,comp.ai,comp.ai.philosophy,sci.philosophy.meta,alt.memetics
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!gatech!newsfeed.internetmci.com!howland.reston.ans.net!ix.netcom.com!netcom.com!jqb
From: jqb@netcom.com (Jim Balter)
Subject: Re: Randomness and free will
Message-ID: <jqbDMoJ3G.7Hq@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <DMBo00.7A@gpu.utcc.utoronto.ca> <4f7oah$ec0@news.cc.ucf.edu>
Distribution: inet
Date: Mon, 12 Feb 1996 20:18:52 GMT
Lines: 89
Sender: jqb@netcom15.netcom.com
Xref: glinda.oz.cs.cmu.edu sci.physics:170865 comp.ai:36971 comp.ai.philosophy:37668 sci.philosophy.meta:24632

In article <4f7oah$ec0@news.cc.ucf.edu>,
Thomas Clarke <clarke@acme.ist.ucf.edu> wrote:
>In article <DMBo00.7A@gpu.utcc.utoronto.ca> pindor@gpu.utcc.utoronto.ca  
>(Andrzej Pindor) writes:
>> In article <4evu83$neg@longwood.cs.ucf.edu>,
>> Tom Clarke <clarke@longwood.cs.ucf.edu> wrote:
>> >yqg023@sunshine.rockwell.com ( Jim Glass ; JF ; GLASS ; x586-0375 ; (W) ;  
>634-000) writes:
>
>> >>In article <823175308.29461@ray.division.co.uk>, ray@division.co.uk (Ray  
>McConnell) writes:
>> >>|> 
>
>> >>A random robot is still a robot.
>
>> >This is an open question.  If the randomness of the robot is
>> >via a random number generator, then the various results about
>> >the capabilities of Turing machines + random number generator
>> >show that no new and interesting behavior should be expected -
>                ^^^^^^^^^^^^^^^^^^^
>> >the Turing machine (robot) would still behavre mechanically
>> >(like a robot).
>
>> Depends on your definition of "new and interesting". Since the alphabet has
>> a finite number of characters, you seem to be implying that no text of final
>> length can contain anything "new and interesting". How come you seem to be
>> reading the usenet postings? 
>
>You got me there.  I seem to be in the grip of an internet addiction,

This is one of the reasons debates here can be interminable -- when someone
is thoroughly rebutted, they rarely acknowledge it.  It appears to me that
Andrzej has shown that your claim about "new and interesting" must be wrong.
If you disagree, what is your argument?

>The issue to me is one of efficiency.  The HLUT will generate
>an intelligent conversation selected from among all the possible
>strings of the alphabet, but it is an impossibly impractical
>machine.
>
>To me it is an empirical question about whether regular, classical
>computer hardware can do the job in a box that can sit beside
>my desk.

This is certainly a significant position and quite worthy of discussion and
exploration.  But it is different from your former statements about the
*theoretical* necessity of QM.  Are you willing to abandon the latter and
stick with the practical issues?  Of course, if so, the philosophy newsgroups
should probably be removed from the list.

><My usual advocacy of the possible relevance of QM deleted> 

One usually advocates a course of action, not a possible relevance.

>> >These are special times in computational research.
>
>> As I have pointed out above, a simple calculation of the number of possible
>> states in the brain (classical), taking just 10^11 0-1 neurons (and the brain
>> is definitely more complex than this) gives you enough states for any purpose
>> you can dream of (and much, much more) _without_ invoking QM.
>
>Computers now have memories with 10^9 bits or more.  They have clock
>rates of hundreds of megahertz, a factor of 10^5 or more faster than 
>the brain. Thus we should not have long to wait before we have a 
>conversation with a computer if you are right.

This would only follow if speed and memory were the only bottlenecks.  But in
fact, software technology is barely at the point of our being able to program
our way out of a paper bag (you can read that literally; perhaps you would
argue that it is easier in the daytime; hoot hoot).  Arguments based upon
failure to achieve some result to date are worthless when we clearly are so
far from the limits.

>I forget exactly
>when its supposed to happen, but according to Moravec computers
>will equal the brain sometime in the next century.  

Neither those who claimed in the 1960's that computers would beat grandmasters
within a decade nor those who claimed that they never would were right.
Arguments based upon either sort of prediction are idiotic.

>Then the issue of the significance of non-classical effects can
>be settled once and for all.

The issue can be settled in favor of the sufficiency of classical effects by
practical results.  The issue can be settled against only by formal proof.
-- 
<J Q B>

