Newsgroups: comp.ai.philosophy,comp.ai,comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!spool.mu.edu!news.clark.edu!netnews.nwnet.net!ns1.nodak.edu!plains.NoDak.edu!vender
From: vender@plains.NoDak.edu (Brad Vender)
Subject: Re: Minsky's new article (was: Roger Penro
Sender: usenet@ns1.nodak.edu (Usenet login)
Message-ID: <Czpt4B.o45@ns1.nodak.edu>
Date: Wed, 23 Nov 1994 10:00:11 GMT
References: <39d8g2$dlm@coli-gate.coli.uni-sb.de> <39eaqk$nn9@cantaloupe.srv.cs.cmu.edu> <CzDqLI.686@cogsci.ed.ac.uk> <jqbCzG3K0.85K@netcom.com>
Nntp-Posting-Host: plains.nodak.edu
Organization: None to speak of.
Lines: 33
Xref: glinda.oz.cs.cmu.edu comp.ai.philosophy:22488 comp.ai:25413 comp.robotics:15590


The problem as I understand it is how to make a machine aware of its current
  state such that it can act on that information.  The simplest explaination
  which I can think of to describe the solution is using a network of
  components or cells.

For the actual physical activity, one set network of cells is set up
  to produce certain outputs for a given set of inputs or sequence
  of inputs (such as a standard sequential digital network).

For the machine's internal awareness, an additional network of cells
  must be set up, with atleast some of those cells possessing
  inputs from the first network (the states of selected cells from the
  physical activity network).
Next, in order to prevent the homonculus problem, instead of building
  another network to be aware of the second network, the second network
  will have to possess cycles (the states of the second network compose
  part of the second networks inputs).
Fortunately, I don't have to even think about how complex a real
  system like this would end up being, or even how complicated it
  would be to design.  However, if someone could build a network
  like this, it leads to a final point....

because this network would probably be designed using some form of
  modern standard techniques and components, it could be simulated
  using a digital computer (even if using entirely analog components),
  meaning that the digital computer doing the simulation was logically
  equivalent (if much slower) to the actual system, and that the
  system was equivalent to a turing machine.

But then again, I'm just a second year engineer....
--Brad
vender@plains.nodak.edu
