home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!cs.utexas.edu!rutgers!igor.rutgers.edu!planchet.rutgers.edu!nanotech
- From: ian@inf.ethz.ch (Ian)
- Newsgroups: sci.nanotech
- Subject: Evolution and nanotech
- Message-ID: <Aug.21.18.42.28.1992.20531@planchet.rutgers.edu>
- Date: 21 Aug 92 22:42:29 GMT
- Sender: nanotech@planchet.rutgers.edu
- Organization: Dept. Biochemistry, University of Bristol, UK
- Lines: 37
- Approved: nanotech@aramis.rutgers.edu
-
- A question which struck me the first time I heard about nanotechnology:
-
- "How do we stop nanomachines from evolving away from their
-
- original purposes ?"
-
- As an illustration consider the case of a nanomachine designed to sit in a bath
- of plastic and produce, for example, telephones. This nanomachine must have
- two distinct functions, i) making telephones and ii) reproducing itself in
- order to replace any defunct machines. The problem arizes with the second
- function because we can not hope to keep the program of a nanomachine entirely
- mutation free (take cosmic ray impact, for example). The classic (modern)
- definition of the requirements for evolution are a) self replication and b)
- mutation so I can't see any way in which we could stop the nanomachine from
- evolving. As a trivial example, a single machine could easily
- develop oncogenic mutations, where the regulation of the reproductive function
- is destroyed and it replicates continually without ever performing its
- telephone making function. From an evolutionary point of view, the mutant
- machine is at an enormous advantage and will almost immediately come to
- dominate the tank, displacing the useful machines!
-
- Now that example is very easily handled. We just throw the defective tank
- into the furnace (nanodissassembler) and start again from a fresh nanomachine.
- Suppose, however that the machine in question had been injected into a patient
- or released into the environment. It seems essential that some mechanism be
- found to prevent this kind of mutation. Does anyone have any ideas ?
-
- I'll save my own ideas until I've seen if anyone flames me :-)
-
- [This is a FAQ and I should really have a good reference set up
- ahead of time. The short form is that we know enough about the
- properties of mutation and evolution that we can build machines
- that do evolve, and we can build machines that do not evolve.
- It would be considered antisocial to do the former...
- Look on planchet.rutgers.edu (standard anonymous FTP) for some
- background material.
- --JoSH]
-