X-Message-Number: 10580
Date: Tue, 13 Oct 1998 11:05:07 -0400 (EDT)
From: Charles Platt <>
Subject: Still More Wishful Thinking

> Message #10573
> From: 
> 

> Opponents see often problems, where problems do not exist. Processing power is
> no problem, as myriads of nanomachines will cooperate.

Excellent argument. Fine! This makes everything much easier. Since we no 
longer need "smart" nanobots--dumb ones can do just as well--the whole 
thing can be simulated right now in software, according to your logic. I 
look forward eagerly to the proof that "cooperating" bots can do the job.

> It has been shown that
> cellular automatons (systems of primitive automatons (e.g. with two states),
> each of which can communicate only with a limited number (e.g. four)
> neighbours) can do everything an universal turing machine can do (if the
> number of machines in the cellular automaton is not limited) 

Glad you added that last little qualifier. Do you have an estimate of the
number of dumb machines required to create, jointly, the processing power
required to solve the horrendously complex problems of 3-D repair work?
Will the number of machines necessary to solve problems involved in
repairing a single neuron fit, say, into a whole human body? 

> The processing power of a system of nanomachines will not depend on the

> processing power of one machine. One bit storage capazity in each machine will

> be enough. The machines can be molecules, if necessary. (Nature uses molecules
> e.g. enzymes for complex tasks.) The processing power of the system will
> depend on the number of machines and this number can be very great.

Enzymes cannot be used to fix freezing damage, so far as I am aware. 

Sorry, but I believe your entire argument is a wild piece of theorizing. 
I have written my own software to create cellular automata (indeed I used to 
sell it via mail order). There is an immense gap between this very limited
numerically controlled behavior, and a huge system that computes cleverly
enough to repair damage afflicting billions of cells. 

Still, as I say, if your system is workable, it can be developed and 
tested right now, using moderate computing power. I look forward to the 
proof of concept.

--CP

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=10580