OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   RE: [xml-dev] More AI Reading

[ Lists Home | Date Index | Thread Index ]

Well that was brainless.  No link; no help. :-/

http://www.ecs.soton.ac.uk/~harnad/Hypermail/Foundations.Cognitive.Science2001/0158.html

Thanks to Joshua Allen for pointing that out.

Just a note on Harnad's commentary and 
we can take this private because this isn't 
particular to XML.  On the other hand, I think 
the symbol grounding problem undergirds many 
of the discussions of WFness and the utility 
of schemas, the need to have a means to 
declare semantics, and why in some applications, 
predicate logic has to be augmented with other 
means.

I think Harnad does not 
want to make a simplifying assumption:  it does not 
matter to the Turing Test that we know if the 
computer simulation "feels".  It matters that it 
can effectively use the symbols we recognize, the signs, 
and provide the illusion.  The Turing Test has 
the problem of the grounding of the tester.  
One can fool some of the people some of the time, 
and empirical evidence with people who are not 
already grounded in computer technology shows 
this.  Often, they react and expect human 
behavior and are only surprised when they 
don't get it.  Competence in the simulation 
cannot be judged independently of the tester.

I believe it is the layering of the simulation 
that must be accounted for.  Harnad makes the 
mistake that for example, love is a thing 
we feel.  Love is a verb.  It is also a thing 
we do.  Emotions are active but not necessarily 
intelligent except insofar as they are sign 
producing actions.   They are not simply 
reflexive, as Gudwin points out, but are 
mediated by experience, by a local grounding in 
episodic memory.  In the Peircean model, 
they are close to the real time interface 
between the external and internal systems 
of the environment and the semiote.   If 
we compare the human to the computer, we 
make a mistake.  If we compare the human 
to a network, we are closer to the physical 
reality of how humans process communications. 

Emotions create the right amount of chaos 
for the semiote to become creative.  Those 
who look for reductionist means to make 
them predicatble will be disappointed.  Those 
who observe them to see if they can produce 
novel behaviors will be delighted.

len




 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS