Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!usenet.ins.cwru.edu!tut.cis.ohio-state.edu!cs.utexas.edu!rutgers!aramis.rutgers.edu!athos.rutgers.edu!nanotech From: alan@oz.nm.paradyne.com (Alan Lovejoy) Newsgroups: sci.nanotech Subject: Re: Drexler on immortality, source of nano books. Message-ID: Date: 23 Mar 90 02:03:12 GMT Sender: nanotech@athos.rutgers.edu Organization: AT&T Paradyne, Largo, Florida Lines: 110 Approved: nanotech@aramis.rutgers.edu In article bfu@ifi.uio.no (Thomas Gramstad) writes: > > >I'm just in the process of reading Eric Drexler's _Engines of Creation_ >and then I remembered that I had seen the word "nanotechnology" >somewhere on USENET.... Yet another group to be followed... Welcome aboard! >That sure is a book with powerful visions. I find it >difficult to assess how much of it is feasible (I'm not an engineer or >technologist, but a biologist (genetics)). Don't miss the forest for the trees. The more specific the technospeculation, the more likely it is to be wrong in some way, and the harder it is to assess (unless it's already known to be (im)possible). But the other side of that coin is that it can be relatively easy to forecast the broad scope of future technology in general terms--with the obvious exception of those things that will use or rely on as-yet undiscovered physical principles. We will acquire ever greater skill at molecular engineering. We will be able to do things which are analogous to those things which existing molecular machines can do. There will be both unforeseen limitations and unforeseen novel capabilities exhibited by future technologies--these are the things that will make us look silly and/or naive to our future selves and to our children. This is the most important theme of "Engines Of Creation." The rest is just window dressing by comparison. >(Sorry, I lost the attribution:) That's ok--now you've found it again: you are quoting me! >>The real problem is the fact that we don't know that a backup of your >>mind is still you (even if you are nothing but information and a >>system of state transition functions). For instance, if we create >>multiple instances of you from a backup, which one is you? Are they >>all you? What does the concept of identity mean in this case? Is >>identity unique, or not? And if not, then why would each one of the >>instances of you all object to being killed? > >>The confusions and contradictions that seem to sprout like weeds when >>one considers this subject suggest to me that at least one of the >>fundamental concepts we use to define/express this problem is flawed, >>inconsistent, meaningless and/or otherwise ill-defined. We don't know >>what we're talking about, at least not fully. > >I think an understanding of our method(s) for concept-formation is >crucial if one is to assess what is possible and not. This is >both an epistemological and a scientific issue. Absolutely. The universe doesn't care what set of conceptual boxes we use to categorize our subjective experience into our internal model of objective reality. A population doesn't change its opinions based on how the statistician/pollster phrases his questions, or on how he decides what group each individual is a member of, or on what groups he decides exist, or on how each group is defined. All those things may certainly affect the results he gets and the conclusions he draws. But they change the reality not at all. We need polling and statistical methods which can measure the underlying reality of a population's opinions. And we need analogous techniques for scientific research which do not depend on how scientific questions are asked, on what conceptual boxes are used to classify the answers, or on what language and semantic system the concepts are defined in. Perhaps all self-consistent conceptual systems are equally valid, but reality can only be approximated to the extent that the number of conceptual systems which are used to think about a problem approaches infinity as a limit. >For example, >even with an accurate understanding of how the mind works, a >simulation of it may still have restrictions or limitations that >the real mind doesn't have (i e don't equate a model with reality). The point has been made elsewhere that a computer simulation of a hydrogen atom, no matter how detailed and/or accurate, can always be easily distinguished from the real thing: just try replacing all the hydrogen atoms in your body with computer simulations to see why this is so. We might call this an imperfect simulation. Imperfectly simulated objects exist in a simulated environment, react to simulated events and interact with other simulated objects. Reality and imperfect simulation do not mix--the real object and its imperfectly simulated twin are not freely interchangeable. However, not ALL simulations have this problem. For instance, computer simulations of other computers can work so well that the only way to tell the difference is to cheat by opening up the box and checking the internal circuits. We might call this a perfect simulation. After some thought about the differences between perfect and imperfect simulations, I have reached the following conclusion: only information and symbolic functions/processes can be perfectly simulated. Perfect simulation is symbolic simulation. Imperfect simulation is non-symbolic simulation. The implications with regard to immortality, identity, artificial intelligence and nanotechnology are obvious. ____"Congress shall have the power to prohibit speech offensive to Congress"____ Alan Lovejoy; alan@pdn; 813-530-2211; AT&T Paradyne: 8550 Ulmerton, Largo, FL. Disclaimer: I do not speak for AT&T Paradyne. They do not speak for me. Mottos: << Many are cold, but few are frozen. >> << Frigido, ergo sum. >>