Path: utzoo!attcan!uunet!husc6!think!ames!lll-tis!oodis01!uplherc!sp7040!obie!wsccs!dharvey
From: dharvey@wsccs.UUCP (David Harvey)
Newsgroups: comp.ai
Subject: Re: AI and Sociology
Summary: Are all forms of programming beneficial?
Message-ID: <577@wsccs.UUCP>
Date: 11 Jun 88 07:00:13 GMT
References: <1033@cresswell.quintus.UUCP> <1301@crete.cs.glasgow.ac.uk>
Lines: 67

In article <1301@crete.cs.glasgow.ac.uk>, gilbert@cs.glasgow.ac.uk (Gilbert Cockton) writes:
> It is quite improper to cut out a territory which deliberately ignores
> others.  In this sense, psychology and sociology are guilty like AI,
> but not nearly so much, as they have territories rather than a
> territory.  Still, the separation of sociology from psychology is
> regrettable, but areas like social psychology and cognitive sociology
> do bridge the two, as do applied areas such as education and management.
> Where are the bridges to "pure" AI?  Answer that if you can.
> 
You are correct in asserting that these are the bridges between
Psychology and Sociology, but my limited observation of people in both
groups is that people in Social Psychology rarely poke their heads into
the Sociology department, and people in Cognitive Sociology rarely
interact with the people in Cognitive Psychology.  The reason I know is
that I have observed them first-hand while getting degrees in Math and
Psychology.  In other words, the bridges are quite superficial, since
the interaction between the two groups is minimal.  In regards to this
situation I am referring to the status quo as it existed at the
University of Utah where I got my degrees and at Brigham Young
University which I visited fairly often.  And in answer to your demands
of AI, perhaps you better take a very good look at how well social
scientists are at answering questions about thinking.  They are making
progress, but it is not in the form of a universal theory, ala Freud.
In other words, they are snipping away at this little idea and that
little paradigm, just like AI researchers are doing.
>
> Again, I challenge AI's rejection of social criticisms of its paradigm.  We
> become what we are through socialisation, not programming (although some
> teaching IS close to programming, especially in mathematics).  Thus a machine
> can never become what we are, because it cannot experience socialisation in the
> same way as a human being.  Thus a machine can never reason like us, as it can
> never absorb its model of reality in a proper social context.  Again, there are
> well documented examples of the effect of social neglect on children.  Machines
> will not suffer in the same way, as they only benefit from programming, and
> not all forms of human company.  Anyone who thinks that programming is social
> interaction is really missing out on something (probably social interaction :-))
>
You obviously have not installed a new operating system on a VAX only to
discover that it has serious bugs.  Down comes the machine to the >>>
prompt and the process of starting the machine up with old OS that
worked begins.  Since the machine does not have feelings (AHA!) it
doesn't care, but it certainly was not beneficial to its performance.
Or a student's program with severe bugs that causes core dumps doesn't
help either.  Then there is the case of our electric news feed being
down for several weeks.  When it finally resumed operation it completely
filled the process table, making it impossible to even sign on as
super-user and do an 'ls'!  The kind of programming that allowed it to
spawn that many child processes is not my idea of something beneficial!
In other words, bad programming is to a certain extent an analog to
social neglect. Running a machine in bad physical conditions and
physically abusing a person are also similar.  Yes, you can create
enough havoc with Death Valley heat to totally kill a computer!
>
> RECOMMENDED READING
> 
> Jerome Bruner on MACOS (Man: A Course of Study), for the reasoning
> behind interdisciplinary education.
>
   ^^^ No qualms with the ideas presented in this book
>
> Skinner's "Beyond Freedom and Dignity" and the collected essays in
> response to it, for an understanding of where behaviourism takes you
> ("pure" AI is neo-behaviourist, it's about little s-r modelling).
>
   ^^^ And I still think his model has lots of holes in it!

dharvey @ WSCCS (David A Harvey)