Newsgroups: comp.ai,comp.robotics,comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornell!travelers.mail.cornell.edu!news.kei.com!bloom-beacon.mit.edu!gatech!howland.reston.ans.net!agate!darkstar.UCSC.EDU!nic.scruz.net!earth.armory.com!rstevew
From: rstevew@armory.com (Richard Steven Walz)
Subject: Re: Minsky's new article
Organization: The Armory
Date: Sat, 10 Dec 1994 15:20:15 GMT
Message-ID: <D0Lp9s.1tI@armory.com>
References: <3c78j8$b44@jetsam.ee.pdx.edu> <3c8esu$jbe@jetsam.ee.pdx.edu> <3c8pur$e53@mp.cs.niu.edu> <3ca2ji$og4@jetsam.ee.pdx.edu>
Sender: news@armory.com (Usenet News)
Nntp-Posting-Host: deepthought.armory.com
Lines: 41
Xref: glinda.oz.cs.cmu.edu comp.ai:25867 comp.robotics:16183 comp.ai.philosophy:23524

In article <3ca2ji$og4@jetsam.ee.pdx.edu>,
Marcus Daniels <marcus@ee.pdx.edu> wrote:
>rickert@cs.niu.edu (Neil Rickert) writes:
>
>>I thought we had agreed to disagree about our conflicting definitions
>>of "free will".
>
>I agree that subjective free will is a valid construct so long
>as actual free will can't be denied.  Denial of free will is
>not reasonable to talk about until the cosmology is fixed.  In this
>case it is, and talking about free will in the everyday practical
>is not meaningful as applied to a computer because the deterministic process
>can't avoid the consequences of itself and its inputs.
>By "everday practical", I mean as applied to blame and responsibility.
>
>It would be vacuous accusation to say to this learning computer 
>"you shouldn't have done that."  It just is what it is, or should I
>say, was.  As free will is a cornerstone to much moral and legal language
>in our culture, for you to say that a intelligent being can't
>avoid its actions is not just a semantic game.  
>
>I can't say that there isn't a non-determinstic representation scheme
>that defies my definition of free will.  You even say that my definition is 
>a tautology, but I don't see how the familiar moral-oughts people
>talk about have any bite without such a scheme.
-----------------------------------------
"Free will" and moral-oughts! I KNEW there was a priest and a couple nuns
in the woodpile!!!;)  Of course you can say to a determined machine that it
shouldn't have done that!! It probably has realized that on its own by that
time, if it makes the mistake at all, as I doubt that you could out-think
it, but such input is not restricted to "entities" with this "magic elixir
of real life" this idiotic "free will". You imagine that a suffiently
programmed learning robot cannot forsee consequences or learn to forsee new
ones!?? How weird and primitive!!! We have seen this sort of learning
routine written for expert system databases already!!! A completely
determined computer can still possess self modifying code, you know, or at
least a field for self-modification of the chains of subroutines and calls!
That's all WE are! You're just terrified of that thought, nothing more
logical will come from you, doubtless! You're frightened and superstitious!
-Steve Walz   rstevew@armory.com

