🚀 go-pugleaf

In <1992Aug26.171151.1898@skynet.uucp> ice@skynet.uucp (Ice) writes:

>In article <9R5NPB5w164w@cybernet.cse.fau.edu> tomh.bbs@cybernet.cse.fau.edu writes:
>>
>>He discusses these issues and give a mathematical proof that simple
>>mechanisms cannot replicate themselves without infinite regress.

>Ummmmmm. I thought John Von Neumann proved that automata of sufficient
>complexity can reproduce themselves. What is Rosen saying that Von
>Neumann did not?

      Can you even give the broadest outlines of how Von Neumann proved this?
I think not. Moreover, whatever the type of logic he utilizied, the only
real proof would be to build one. Since size is not really relevent to the
design process, I would like to know ballpark how many parts such a machine
would have? Don't forget, the more parts it has, the more parts it must have
to build those parts -- since all the parts cannot be the same.

      The argument is that it is _very_ complicated, but not impossible.
But it seems to me we should at least be able to characterize this complexity
as deriving from a certain problem. I invite someone to try it, and yet
stay clear of infinite regression.


>Ice()


      John E. Snodgrass
      -----------------
      snodgras@crash.cts.com