Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uflorida!haven!adm!cmcl2!yale!engelson From: engelson@cs.yale.edu (Sean Engelson) Newsgroups: comp.ai Subject: Re: Question on Chinese Room Argument Message-ID: <52507@yale-celray.yale.UUCP> Date: 2 Mar 89 18:26:38 GMT References: <4298@pt.cs.cmu.edu> <8174@netnews.upenn.edu> <51123@yale-celray.yale.UUCP> <2483@crete.cs.glasgow.ac.uk> Sender: root@yale.UUCP Reply-To: engelson@cs.yale.edu (Sean Engelson) Organization: Computer Science, Yale University, New Haven, CT 06520-2158 Lines: 58 In-reply-to: gilbert@cs.glasgow.ac.uk (Gilbert Cockton) In article <2483@crete.cs.glasgow.ac.uk>, gilbert@cs (Gilbert Cockton) writes: >In article <51123@yale-celray.yale.UUCP> engelson@cs.yale.edu (Sean Engelson) writes: >>(Searle + rules) understanding of Chinese? It seems that to >>demonstrate or refute the position of understanding being demonstrable >>purely through I/O behavior, one must have an effective definition of >>understanding. By effective I mean one that does not beg the >>question, i.e. by defining understanding to be symbol-processing, or >>conversely, to be that which humans do. > >Sorry, but your constraints are a little weird. Understanding *IS* >what humans do. >It *MAY* involve symbol processing. What question is begged? What is >an effectiver >definition? What I meant was rather "that which _only_ humans do", i.e. a priori ruling out any form of non-human understanding. >Look how well physicists manage with "force", "charge", "gravity". >You cannot ask commentators on humanity for "definitions" that are >any less (fast >and) loose than those used by commentators on nature. But a physicist can give me a simple and effective procedure by which I can measure the charge of a body, or the force of gravity. I have seen no such procedure of criterion for recognising understanding, other than I/O equivalence with that which we call understanding in humans. Under that criterion, the Chinese Room understands. >The question for the strong AI brigade is: > >"Given the normal usage of understanding, what grounds are there for >attributing it >to computers, and why bother anyway" > >While we're at it, what about those halucinogenic thermostats with beliefs. >Whatever it was, don't eat it again :-) The normal usage of understanding is that if (a) someone says they understand, and (b) they act as if they do, then they understand. Unless you have a theory of understanding that rules out physical symbol systems (and neural nets are symbol systems too!), then I see no reason not to attribute understanding to computers. Why bother anyway? Excellent question. I see no real purpose in it, except that the anti-attribution thereof is used as a criticism of AI, incorrectly. If it looks like a duck, and it acts like a duck, and it quacks like a duck, then what does it matter if it understands in an epistomological sense or not? ---------------------------------------------------------------------- Sean Philip Engelson, Gradual Student Who is he that desires life, Yale Department of Computer Science Wishing many happy days? Box 2158 Yale Station Curb your tongue from evil, New Haven, CT 06520 And your lips from speaking (203) 432-1239 falsehood. ---------------------------------------------------------------------- Nondeterminism means never having to say you're wrong.