Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucsd!rutgers!aramis.rutgers.edu!athos.rutgers.edu!nanotech From: sean@aipna.edinburgh.ac.uk (Sean Matthews) Newsgroups: sci.nanotech Subject: Re: Rendevous with Rama Message-ID: Date: 20 Sep 89 03:03:35 GMT Sender: nanotech@athos.rutgers.edu Organization: Dept of AI, Edinburgh University, UK. Lines: 41 Approved: nanotech@aramis.rutgers.edu In article I write: >P.S. should `nanotec(h)ology' be spelled with or without the `h'? > >[I assumed you meant "nanotechnology" and corrected the spelling above. > --JoSH] No, the spelling was deliberate, I intended a portmanteau on `nanotechnology' and `ecology'; i.e., an artificial system that looks like a natural ecology. It seems to me that, in some ways, the Rama system gives a good picture of some parts of a large system built using those techniques. All the `machines/lifeforms' in Rama give the impression of being entirely artificial, but organised on the lines of `natural' life; I think that any such large machine we might build would tend, since we would want all the useful properties that life (like self organisation) has in such a machine, to give the same impression of rationally reverse engineered life. (and as we all know, it is cheaper to reverse engineer proven principles than it is to start anew). It would be much easier to produce a robot like a Raman from a tank of chemicals that it would be to produce a robot like a Unimate. The reason that I think this is the case is that life is `engineered' on self similar principles that work from the very bottom up (most people will have seen the example of the fern generated from a fractal), which is how the sort of manufacturing technolgy that Drexler et al. envision would work. Current engineering is based on the idea of everything just coming together as a whole, since it can rely on the godlike perspective of the person or machine that is puting it together. Getting such a system to work bottom up would be *difficult*. Sean