Re: should we avoid attributing mental states to AI mechanisms ?

From: David Longley (David_at_longley.demon.co.uk)
Date: 06/06/04 

Date: Sun, 6 Jun 2004 18:55:13 +0100

In article <vQHwc.9884$HG.1870@attbi_s53>, patty
<pattyNO@SPAMicyberspace.net> writes
>David Longley wrote:
>> In article <Xrrwc.10849$Sw.9444@attbi_s51>, patty
>><pattyNO@SPAMicyberspace.net> writes
>>
>>> JXStern wrote:
>>>
>>>> On Fri, 04 Jun 2004 18:00:36 GMT, patty <pattyNO@SPAMicyberspace.net>
>>>> wrote:
>>>>
>>>>> Incidentally do not attribute my words about inventing worlds to
>>>>>Pat Hayes, they are but my flaky paraphrase of an understanding
>>>>>that i picked up in my meanderings. In fact i cannot even find
>>>>>where Pat said what i said he did .. all that i can find is where
>>>>>other people attributed that saying to him ...
>>>>
>>>> The unicorns bit is good, but I like the "inventing world" phrase even
>>>> better. I don't care who said it, but would like to see some full
>>>> article or book where the idea is developed or used. Again, if you
>>>> have a reference for either or both of those, I'd appreciate it - I
>>>> gather you didn't refer me to w3 or common logic for those.
>>>>
>>>
>>> Well i don't have a reference, but i still claim that this is
>>>nothing new, and is just a standard way of thinking when doing formal
>>>semantics. Perhaps you will find your reference somewhere in a paper
>>>on formal semantics. The RDF Semantics document which i referenced
>>>is an example of this ... just note how Pat defined the
>>>interpretation and how he drew the diagram.
>>>
>>> I think that when we try to understand what a person is saying we
>>>frequently concoct a imaginary world which contains the assumptions
>>>which we have attributed to the speaker and comprehend his words in
>>>relationship to that world, rather than our own. However those of us
>>>who do not use that strategy more frequently can make no sense
>>>whatsoever of what others are talking about.
>>>
>>>> (the idea of the common logic thing, that's another topic entirely!
>>>> have you ever read Lewis Carroll's "Achilles and the Tortoise",
>>>> published in the philosophy journal Mind in 1895?)
>>>> http://www.mathacademy.com/pr/prime/articles/carroll/index.asp
>>>> (btw, I disagree entirely with the analysis given on this page, just
>>>> stick to the Carroll stuff and work the problem yourself, I suggest)
>>>>
>>>
>>> Well in regards to the last paragraph where he says ...
>>>
>>> ""
>>> The formalist solution, while effective, has its own philosophical
>>>drawbacks. Not the least of these is that, by reducing logic to
>>>uninterpreted symbols, all semantic content is removed from the
>>>conclusions of formal logic. In other words, what we would ordinarily
>>>consider meaning is lost. How to restore meaning to systems of
>>>inference while still avoiding difficulties such as Carroll’s
>>>Paradox remains a thorny question for philosophers of mathematics.
>>> ""
>>>
>>> I would just like to note that when the semantic content is restored
>>>to the symbols, some of the formal assumptions (usually the law of
>>>the excluded middle) become invalid. Which, i claim, is why
>>>classical logic rarely works in real life.
>>>
>>> patty
>> I don't know what this "real life" is that *you* keep alluding to.
>>>From the sound of it, I'd go as far as saying that you don't know
>>what you're talking about, and that you have no hope of ever telling
>>us either if you persist in talking this way.
>
>Real life is what humans experience between birth and death. It is the
>events that happen inside their "experiential bubble".
>
><http://home.tiscali.nl/boynalechmipo/index.htm>
>
>> Our economic and general physical well-being is critically dependent
>>on the extensional stance. I don't know what you mean by "classic
>>logic" either, but if you mean modern, post-Fregian logic, we depend
>>on it. To a very large extent "classic" logic runs our "real life",
>>it certainly runs the Internet!
>
>I take classical logic to be the kind of logic discussed on the page
>about Carroll’s Paradox, see above. As noted on that page, the
>symbols that are processed are devoid of the content that we find in
>real signs. Yet when we try to apply classical logic we must still
>assume that a strong identity on these symbols are possible and use
>that to exclude possibilities (LEM) which actually exist in the world
>around us. Fact is, this logic only works in toy domains and in
>academic games. It has been hyped in our culture far in excess of any
>hype associated with AI. Were it never invented by man, our science
>and technology would be just as advanced as they are today.
>
>Our world runs on physical mechanism and we depend on them. Computers,
>for the most part, run on if-then statements, not the modus ponens of
>classical logic. In many cases we utilize statistical analysis. These
>statical heuristics do not gain their utility from anything that could
>be called "extensional" using your definition of that term.
>
>patty

That's all bull***. If you can't see why, you are deluded. 

-- 
David Longley