Newsgroups: comp.ai
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!news.duke.edu!godot.cc.duq.edu!newsfeed.pitt.edu!dsinc!ub!galileo.cc.rochester.edu!prodigal.psych.rochester.edu!stevens
From: stevens@prodigal.psych.rochester.edu (Greg Stevens)
Subject: Re: "Your autonomous agents have no clothes!"
Message-ID: <1995Mar16.235159.1671@galileo.cc.rochester.edu>
Sender: news@galileo.cc.rochester.edu
Nntp-Posting-Host: prodigal.psych.rochester.edu
Organization: University of Rochester - Rochester, New York
References: <3jt3f4$4ej@Mars.mcs.com> <NAGENDRA.95Mar12143152@mesmer.cs.umass.edu> 	<3k9qiv$ab6@Venus.mcs.com> <NAGENDRA.95Mar16130159@mesmer.cs.umass.edu>
Date: Thu, 16 Mar 95 23:51:59 GMT
Lines: 29

In <NAGENDRA.95Mar16130159@mesmer.cs.umass.edu> nagendra@mesmer.cs.umass.edu (MARAM V. NAGENDRAPRASAD) writes:
>On 16 Mar 1995 10:55:59 -0600, jorn@MCS.COM (Jorn Barger) said:

>> If I understand the latter school (which is new to me), the important
>> realization is that if you have two or more CPUs trying to work
>> together at almost any level, the sorts of difficulties that arise
>> will be *deeply* parallel to social-psychological phenomena like
>> negotiation, resource-allocation, priorities, and authority?

>I think you got the point (the same cannot be said of a lot of people who
>see our work for the first time). At a higher level (as versus distributed
>computing) issues like partial information, uncertainty etc lead to
>satisficing problem solving (just as in social systems) and much of
>our study involves gaining insight into various trade-offs in such systems.    

It would be interesting to me to know what methods other than satisficing
problem solving (if indeed you are right when you use this term) have
been used in this application, considering that in the psych-social sciences
there is very little agreement as to whether "satisficing" (as Herbert Simon
originally used the term) is what people ddo in the presence of uncertainty
and computational limitations, rather than any of the various other
methods of computation-reduction, including Tversky and Kahnemann's own
Prospect Theory involving the heuristics they theorized from their
empirical work.

Greg Stevens

stevens@prodigal.psych.rochester.edu

