NICK@AI.AI.MIT.EDU (Nick Papadakis) (06/02/88)
From: Gilbert Cockton <gilbert%cs.glasgow.ac.uk@NSS.Cs.Ucl.AC.UK> Date: Mon, 30 May 88 07:13 EDT To: AIList@ai.ai.mit.edu Subject: Immortality, Home-made BDI states and Systems Theory (3 snippets) In article <8805250631.AA09406@BLOOM-BEACON.MIT.EDU> >Step (4) gives us virtual immortality, since whenever our current >intelligence-carrying hardware (human body? computer? etc.) is about to >give up (because of a disease, old age ...) we can transfer the >intelligence to another piece of hardware. there are some more delicate >problems here, but you get the idea. Historically, I think there's going to be something in this. There is no doubt that we can embody in a computer program something that we would not sensibly embody in a book. In this sense, computers are going to alter what we can pass on from one generation to another. But there are also similarities with books. Books get out of date, so will computer programs. As I've said before, we've got one hell of a maintenance problem with large knowledge-based programs. Is it really going to be more economical than people? See the latest figures on automation in the car industry, where training costs are going through the roof as robots move from single functions to programmable tasks. GM's most heavily automated plant (Hamtramck, Michigan) is less productive than a much less automated one at Fremont Ca. (Economist, 21/5/88, pp103-104). In art. <8805250631.AA09382@BLOOM-BEACON.MIT.EDU> > >Let my current beliefs, desires, and intentions be called my BDI state. These change. Have you no responsibility for the way they change? Do you just wake up one morning a different person, or are you consciously involved in major changes of perspective? Or do you never change? In article <19880527050240.9.NICK@MACH.AI.MIT.EDU> >Gilbert Cockton: Even one reference to a critique of Systems Theory would be >helpful if it includes a bibliography. I recommended the work of Anthony Giddens (Kings College, Cambridge). There are sections on systems theory in his "Studies in Social and Political Theory" (either Hutchinson or Macmillan or Polity Press, can't remember which). A book which didn't impress me a long time ago was Apple's "Ideology and Education" or something like that. He's an American Marxist, but I remember references to critiques of systems theory in between his polemic. I'll try to find a full reference. Systems theory is valuable compared to classical science. To me, systems theory and simulation as a scientific method go hand in hand. It falls down in its overuse of biological concepts (which with mathematics represent the two scientific influences on many post-war approaches to humanity. Sociobiological game theory, ugh!) Another useful book is David Harvey's "Science, Ideology and Human Geography" (Longman?), which followed his systems theory/postivist "Explanation in Geography". You'll see both sides of systems theory in his work. Finally, I am surprised at the response to my original comments on free will and AI. The point is still being missed that our current society needs free will, whether or not it can be established philosophically that free will exists or not. But I have changed my mind about "AI's" concern about the issue, both in the orderliness of John McCarthy's representation of a 1969 paper (missed it due to starting secondary school :-)), and in Drew McDermott's awareness of the importance of the issue and its relation to modern science and dualism, plus all the other traffic on the issue. I only wish AI theorists could get to grips with the socialisation question as well, and understand more sympathetically why dualism persists (by law in the case of the UK school curriculum). Hope you're enjoying all this as much as I am :-) Gilbert.