Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!mit-eddie!media-lab!minsky From: minsky@media-lab.MEDIA.MIT.EDU (Marvin Minsky) Newsgroups: comp.ai Subject: Re: Types of AI thought Keywords: Chris Malcolm says, Message-ID: <3520@media-lab.MEDIA.MIT.EDU> Date: 2 Oct 90 02:38:03 GMT References: <3095@aipna.ed.ac.uk> Reply-To: minsky@media-lab.media.mit.edu (Marvin Minsky) Organization: MIT Media Lab, Cambridge MA Lines: 13 > Much too simple. But among those who are trying to understand > anything new the first stage is always the realisation that there > are two kinds -- the first categorisation. An essential first > stage. But only the first: beware of those supposed experts who > tell you that there are two kinds of anything! Hear, Hear! These are what we call dumb-bell theories, and they are indeed good starts. But beware of those who have gotten no further. If any such dichotomy stays around more than a short time, then it was proabbly a bad start, and you shold consider abandoning it. Certainly "soft" and "hard" AI make no sense. I have seen no progress from Piaget's great 'assimilation" vs. 'accomodation' distinction. David Marr's well-known 3-types don't appeal to me, either.