Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!EU.net!CERN.ch!dxal18.cern.ch!hallam
From: hallam@dxal18.cern.ch (Phillip M. Hallam-Baker)
Subject: Re: Books on philosophy of AI
X-Nntp-Posting-Host: dxal18.cern.ch
Message-ID: <D3DLF2.34L@news.cern.ch>
Sender: news@news.cern.ch (USENET News System)
Reply-To: hallam@dxal18.cern.ch
Organization: Wot!!! Me ????
References:  <3gl39k$1op@odin.diku.dk>
Date: Thu, 2 Feb 1995 13:57:02 GMT
Lines: 51


In article <3gl39k$1op@odin.diku.dk>, kash@diku.dk (Kasper Hornb{k) writes:

|>Hi.
|>I'm doing an essay of philosophical problems in AI. Does anybody
|>know of some books or papers regarding this topic. If so please
|>e-mail me or post to this group.

Oxford University Press has published an anthology (author Boden) of
seminal AI philosophy papers. It is called the Phillosopy of Artificial
Intelligence. V. good IMHO.

Just don't spend too long getting wound up by Searle's chinese room :-)
I think the true interest of the problem is not how to squish it but how
to squish it in a novel way without writing a paper the length of a novel.
Its a pretty silly argument and he employs circular reasoning in a very
annoying fashion "If I don't understand chinese does it make sense to say
that myself plus a few bits of paper can understand it?". The bits of paper
being precisely the computer program which he is fighting against. Huh!


I would also look at Neurophilosophy by P. Churchland - its more for those
people who think we should be building artificial brains rather than 
artificial intelligence. The former may be an interesting academic challenge
but the result is not as usefull as the second. A grad student who comes
up with a replica of Newt Gingrich's thought process would not impress me
as much as someone who came up with a system that could analyse say
international politics and come up with interesting, reasoned and novel
propositions.

Interesting:	Proposition must have some relevance to a question at issue

Reasoned:	The proposition must be justifiable in terms of the data.

Novel:		The proposition must be a non trivial deduction from the data
		set such that a human would not make it immediately.


The "Newt Gingrich" analyser would have some limited use however, one might
use a trivial deduction system to analyse data from covert agents whose 
identities must be protected.


Sorry about the ad-hominen attack on the grate leader. But I've just read
yet another "say the first think that enters head" statement from him.


--
Phillip M. Hallam-Baker

Not Speaking for anyone else.
