[extropy-chat] Why no assembler design?

[Eugen Leitl]

On Wed, Nov 19, 2003 at 09:16:35AM -0500, Dan Clemmensen wrote:

> It doen't matter what I'm trying to build. This is the preliminary 
> bootstrap assembler. A hand-cranked wooden lathe, which we then use 
> to build a brass lathe, which we then use to build a steel lathe, 
> which we then use to build a rifle.

I realize that. I was just trying to be cute on the cheap.
(Btw, you're difficult to quote due to overlong lines).

> I suspect that there are many potential approaches to building the first primitive assembler, and furthermore I suspect that you, Hal, Robert, and Chris, at least, are more up to date than I am. That's not the point. The point is that we should focus on building that very first primitive assembler, and more specifically on finding a design and a construction method that is achievable starting from today's base. It does not matter how crude, or slow, or expensive the bootstrap assembler is, as long as it can build something better than itself. We only need one of them. (More specifically, it only needs to succeed once.)

My point is that there's a rich set of bootstrap scenarios,
with very uncertain outcomes given our current knowledge, 
and it is hence way too early to hedge our bets.

We should try all of them, and then focus on whatever
one starts to pan out first. Luckily, you need only a good
lab for one approach, and we've got plenty of labs capable.
Difficult to make them interested that early in the game,
though.

Remember, the assembler has become synonymous with cargo
cult in the mainstream science (not press, that's something
different). People will actively discredit you and try to
get your funds cut if you proclaim assemler is your design
target.

> My example is very sketchy and not well analyzed. IT was more intended as an example of a progression. There are two critera for the bootstrap assembler: It must be buildable without using an assembler, and it must be able to build the next assembler in a progression. (We can and should state both criteria with more rigor.) Just as my "primitive nanotube-based assembler" may be a poor choice to satisfy the buildability criterion, my progression may be a poor example of the correct progression from the nanotube assembler.
> 
> The point is to agree that the bootstrap assembler is the best goal to concentrate on now, and that the two critera are a good way to focue the effort.

We should focus on nanotechnology in general, not on the assembler alone,
and not on a specific flavour of bootstrap. Imho, of course.

> 
> Drexler spend most of "Nanosystems" exploring Diamondoid nanotechnology in depth, and then sketched out some bootstraps. We have some new bootstrap approaches to explore that may be superior. This is where the effort needs to be spent.
> 
> I've been away from this for the last five years. I apologize if I am re-hashing old ground.

I'm not sure just discussing this stuff matters. 
Reality tends to become rather normative. 

No one will even remember this group. All this work will be wasted,
if we just keep talking. It's okay if we set out to generate hot
air, but I presume a number of folks here are not, and might get
unduly disappointed. 

-- Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20031119/4f97e078/attachment.bin>