Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!usc!rutgers!aramis.rutgers.edu!athos.rutgers.edu!nanotech From: kerce@nu.cs.fsu.edu (Kingsley F. Kerce) Newsgroups: sci.nanotech Subject: Competition is cognition (was Re: Some problems of super-intelligence) Message-ID: Date: 13 Dec 90 21:31:55 GMT Sender: nanotech@athos.rutgers.edu Organization: Florida State University Computer Science Lines: 48 Approved: nanotech@aramis.rutgers.edu In article autodesk!peb@uunet.uu.net (Paul Baclaski) writes: >In article , >dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: >> ...Suppose a theoretical limit exists to the maximum amount of >> intelligence that can exist in one coherent entity, before the >> subparts become so intelligent that they create their own >> independent agendas and rebel? > >Consider human organizations--the larger the organization, the >more bureaucracy occurs. Minksky proposes in his Society of >Mind (and in the epilogue of the new edition of Perceptrons >(which I highly recommend for some critical analysis of connectionism)) >that subparts would get gross overviews of what other subparts are >up to. The more subparts and the higher the bandwidth, the more >difficult this will be. Competition between subparts resulting in whole benefit is commonplace in nature. In fact, neural development is a very selective process in which neurons vie. This was first proposed in the late 1800s. In developing nervous systems, cells are overproduced and neuronal death plays a major role in the final neural pattern. For example, covering a newborn kitten's left eye, uncovering it after three months, and applying a light stimulus to that eye results in no activity in the visual cortex. All inputs to the visual cortex come from the right eye. The kitten is born with all the appropriate neuronal connections in its visual system, however, if one eye is covered all the connections are taken over by the other eye. The immune system is selective. Subparts of this protective system do not dynamically adjust their structure to fight invaders. Instead, the subparts which are indeed effective against intrusion--by virtue of their structure--will increase in quantity. In theory, selection occurs in groups of neurons (see G. Edelman's Theory of Neuronal Group Selection). At a higher level (see R. Brook's subsumption architecture for robot control) behaviors may compete for expression. In many circumstances, competition is healthy. -- Kingsley Kerce kerce@nu.cs.fsu.edu Dept of Computer Science B-173 FSU, Tallahassee, FL, 32306-4019 Work Phone: (904)644-8562