Re: Constraints on the "singularity"

Dan Clemmensen (Dan@Clemmensen.ShireNet.com)
Mon, 13 Oct 1997 19:00:31 -0400


Ramez Naam (Exchange) wrote:
>
>
> Possibly you're right, and the superintelligences are incomprehensible.
> However it seems that we should be able to at least generate some
> quantitative boundaries on their capabilities given our deepest current
> understanding of the laws of physics and mathematics.
>
> E.g.:
[Snipped good examples]
The Key is that these bounds are based on our deepest understanding,
not in the understanding of the superintelligences. If thg SIs can
figure out way around the constraint we see then all bets are off.
I personally don't concern myself with new laws of physics until
I see some compelling experimental evidence, but I'm not an SI, so
I cannot begin to design the relevant experiments. I'll restrict myself
to helping to build the SI. What new physics might our hypothetical
SI find? well consult any good scienc fiction collection. The current
universe may not be the only accessible universe, as a sufficient
example.

> > > Chaotic Computability Constraints: The most ambitious nanotech
>
> My understanding may be off here, but let me put forth the reason I see
> chaotic computability entering into the picture with nanotech:
>
[SNIP of simple STM]
> PARALLEL NANOTECH
> Let's use "Parallel Nanotech" to describe the use of a swarm of
> non-reproducing assemblers to transform the source material into the
> desired object(s). Possibly this can be handled without worrying about
> chaos, though I'm not sure. The question here is what control system is
> used to distribute instructions to the appropriate assembler at the
> appropriate time. There are several options in this area, divisible
> into two groups:
>
> Group 1) Control systems requiring central planning and coordination of
> the assemblers. Ie, with electromagnetic signals of some sort I could
> communicate to each nanite and instruct it in what to do. Or I could
> "draw" the shape of the object I wanted and have different classes of
> nanites respond to different signals (which correspond to different
> areas). This requires external planning, analysis of the source
> material, and either precise regulation of the source quality and
> structure, or on-the-fly recalibration of the output material. This
> type of control seems to avoid the chaotic computability problems but is
> a much less flexible nanotech than is popularly conceived by nanotech
> proponents.
>
[SNIP of much more complex systems requireing serious autonomy]

A distributed system can consistently build a fixed design to atomic
precision if the design is hierarchical. Such a strategy may not be
consistent with the wilder nanotech scenarios, but i should more than
suffice to build replacements for all of the hard goods that comnprise
the current basis of the economy, and to build truly massive computers.
IMO this is far more than enough to obsolete the current economic
system and provide the computation basis for the SI. Stochastic
assembly and other non-deterministic systems are unnecessary even
though they may be interesting.

A hierarchical system is describes at some length in Nanosystems.
To summarize: specify a bunch of building blocks, and assemblers or
"factories" to build them from atoms. Also specify next-level
assemblers and/or factories to build next-level building blocks.
Go as many levels as necessary.

Sure, such a system is limited, but who cares? It's effectrively
a great deal less limited than any current factory.