[extropy-chat] self replicating machine ....

[Dan Clemmensen]

Adrian Tymes wrote:

>--- Dan Clemmensen <dgc at cox.net> wrote:
>  
>
>>I changed my mind. Rather than using a technology based on X-Y
>>plotters that
>>create plots for humans, let's look at technology that is
>>specifically
>>designed for optimal self-replication. We want to use simple
>>substrate 
>>materials
>>and a simple design. My guess is
>>    
>>
>
>The problem is, using already-existing stuff means you don't have to
>make (potentially wildly inaccurate and proof-defeating) guesses.
>
>So, I'd suggest sticking to things that do exist.  Designing self-rep
>by some - any - means is hard enough, without restricting yourself to
>novel means.
>_______________________________________________
>  
>
Adrian, I have spent some considerable time looking at SPM designs. 
These designs can
achieve approximately 10nm resolution using  piezo-electric elements. I 
picked
10 micrometer resolution as an extremely conservative goal. This is not 
some aggressive
new technology. The only reason that this technology is not in use today 
is that
it is obsolete: given the existence of commercially-available 
piezo-electric elements,
there is no reason to use electromagnetic servos to achieve this accuracy.

I choose to use electromachanical servos because they have a much simpler
substrate. the substrate for a servo is essentially the same as the 
substrate for an electric
motor;. By contrast, the substrate for a piezo involves complex 
extraction and synthesis
systems.