home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!newsserver.jvnc.net!rutgers!igor.rutgers.edu!planchet.rutgers.edu!nanotech
- From: cuhes@csv.warwick.ac.uk (Malcolm McMahon)
- Newsgroups: sci.nanotech
- Subject: Re: The Industirial Egg or Why Nano? (was Surviving)
- Message-ID: <Jan.22.22.37.12.1993.11589@planchet.rutgers.edu>
- Date: 23 Jan 93 03:37:13 GMT
- Sender: nanotech@planchet.rutgers.edu
- Organization: Computing Services, University of Warwick, UK
- Lines: 50
- Approved: nanotech@aramis.rutgers.edu
-
- First off the "Industrial Egg" idea as presented is actually a little
- impracticable in a world were the nearest source for some of the
- necessary raw materials might be on another continent and somebody
- else's property. Such a machine would need to start life with a hefty
- bank balance and a small army of andriod lawyers so we'll assume, for
- the sake of argument, that you deliver it's raw materials as is the
- intention with nanomachines anyway.
-
- Building a replicator using bulk technology is possible in
- principle, just as it is possible in principle to build a general
- purpose computer out of electromagnetic relays. You could probably make
- such a computer with turn of the century technology (the last century
- that is!) and it could be made to work. A machine of complexity
- compatible with, say, an 8 bit games machine would probably be about
- the size of a house and might reliably manage something of the order of
- four or five instructions per minute, maybe as high as ten.
-
- The central point is that, all other things being equal, the time
- it takes a device to carry out an action is roughly proportional to the
- linear scale of the device. This means that the reproduction time for
- your bulk technology replicator would be of the order of a billion
- times longer than that of a nanomachine. To put it simply by the time
- your first car came off the production line a nanomachine could
- probably have covered the land surface of the Earth about ten cars deep
- (which is the kind of reason that the technology is so dangerous).
-
- Of course a large machine can handle larger workpieces but the
- point with replicators is that the replication speed has an exponential
- influence whereas the the size of the workpiece has only a linear
- effect. This means that to the time for nanotechnology to build an
- artifact varies only logarithically with the mass of the artifact. Try
- to do the same thing with large scale technology and not only will it
- take years to get going you'll rather quickly run out of planet (in
- space it might make more sense).
-
- Using evolution to develop replicators is an interesting but
- highly risky enterprise. Frankly it's the kind of thing I'd rather
- watch from a safe distance, say from Mars. You'd have to be extremely
- thorough about how you imposed your success criterior or you'd be in
- danger of finding that the "fittest" replicator turned out to be the
- dreaded "grey goo", a replicator which converted everything it
- encountered (including the experimental containment) into copies of
- itself. A safer, and in many ways superior way to try this would be in
- computer simulation.
-
- On the subject of the transition from industrial society; yes I
- think it has already started. Only in Japan is there any recognition of
- this. We aint seen nothing yet!
-
- Malcolm McMahon
-