Subject: [extropy-chat] Social Implications of Nanotech

[Dan Clemmensen]

Robin Hanson wrote:

> I am struggling to clarify and identify the differing assumptions that
> different people are making, and place them in economic terms to support
> economic analysis.  *I'M OPEN TO SUGGESTIONS*, though I'm running out of
> time for this round.  How would you describe the differing assumptions,
> making them as explicit as possible, and trying not to assume everyone
> who disagree with you is an idiot?
>
I think that, as you said most of us start with the Basic MNT assumption 
that design and build to atomic precision is possible.

I feel that most people assume that the contol system for a 
self-replicator is a solved problem. I cannot agree with this, and I 
think it is the crux of the issue. We need both the low-level technology 
of MNT, and the control system. I think we will get the control system 
before we get MNT, but the control system is a difficult problem to 
solve. Has anyone demonstrated a control system for a self-replicator in 
an environment of complexity equivalent to an MNT "factory?" I think not.

It's a small matter of programming.    Or perhaps a program for small 
matter.

Here are my assunptions for a bootstrapped MNT technology:
1) It must be possible to create a suitable raw environment using 
macro-scale techniques. An example  of a suitable raw environment might 
be  a glass bottle filled with 99% pure Argon and 1% carbon dioxide with 
a 1 cm^2 HOPG substrate and a  laser for power and control, and a 
continuous gas feed.
2) It must be possible to introduce a MNT assembler into the environment 
using macro-scale techniques.
3) It must be possible to build the first assembler "by hand" using 
macro-scale techniques
4) It must be possible to direct the activities of the assembler from an 
external control system.

Items 3 and 4 are just a bit difficult :-)

I'm fairly sure that we can design control system hardware entirely from 
parts that can be built with an assembler chain (below) using design 
tools and techniques that we can develop with today's tools. I am not 
sure that we can design the software, but if we can, then the system as 
a whole is self-replicating.   Self-replication requires that the 
assembler be able to build copies of itself from CO2, and that asemblers 
are able to build small macro-scale assemblers and parts for still 
larger assemblers, forming a chain of progressivly larger assemblers.

The bootstrap control system hardware can clearly be a macro-scale computer.

So we are back to Drexler's argument: we only need to build one MNT 
assembler "by hand," and write some software.

Just these two "little" problems between us and the end of the world as 
we know it.