The Spike, nanotech, and a future scenario

Mitchell Porter (mitch@thehub.com.au)
Fri, 3 Oct 1997 22:36:40 +1000 (EST)


This evening I finally got hold of a copy of Damien Broderick's
_The Spike_ outside of the confines of a bookstore. I'm pleased
with the breadth of its coverage of transhumanist ideas. I'm also
glad that something has been written about the online >H culture
of '96-'97, something beyond the web essays and list archives
that will no doubt remain suspended in cyberspace. The book has
some of the feel, for me, that being a participant online has had;
and I think the feel of things "here" will change as transhumanism
obtains a higher profile, and as new concepts of the future emerge.
Anyway, I enjoyed it.

Since my US trip, I've become much more involved with nanotech -
I studied _Nanosystems_ while riding the Greyhound buses out to
Dallas and Zyvex, and I formed the conclusion that it should
be incredibly easy to put together complete assembler designs
(at the component-level, not the atomic-level) from what's in
that book. Since my return to Australia, I found a web paper by
Ralph Merkle specifying, so he claims, a set of molecular tools
capable of synthesizing "all the rigid hydrocarbons", which would
include any customized diamond shape (diamond*oid* might involve
elements other than carbon and hydrogen, but pure diamond seems
enough to realize the whole gamut of nanomechanical devices).
Also since my return, Zyvex have finally obtained the atomic
probes that they were waiting for when I visited, so they've
started the real lab work.

All in all, nanotech looks extremely close to me. And the arrival
of MNT, I think, would invalidate the Moore's-Law type of projection
which places Singularity in 2035 or 2025, since Moore's Law
is ultimately based on the rate of improvement of semiconductor
technology. Nanotech rod logic would provide capabilities equal to
those which a Moore's-Law projection might predict for
the late 2010s. But I can imagine a Merkle-type "hydrocarbon
assembler", using buckytubes for the bodies of the molecular tools,
being made by 2005 or even by 2000.

So I'm spending a little time trying to be a good Yudkowskian
Singularitarian, aiming for Singularity now by doing my part
to hasten a complete assembler design, and spending a little
time wondering about the wisdom of hastening the arrival of a
technology which would allow one to make a planet-killing device
smaller than a bacterium - *especially* about the wisdom of
working on it in a public-domain collaboration (like
http://nanocomputer.org).

And I'm spending a little time on scenario generation. An example:

After the first assemblers built, there is indeed a
period of time in which "everything goes haywire": First
politically and psychologically haywire, as the powers that be all
seek a stake, and as the sort of ideas discussed in _Engines
of Creation_ are in the news media every day, accompanying
the latest reports on nanotechnological progress. Then
physically and economically haywire, as matter compilers spread
throgh society and nanomachines become part of our physical
environment (e.g. in the form of grey goo and countermeasures!).
We might see traffic into space increase by orders of magnitude
and a vast increase in the depth and breadth of net connectivity
(imagine everyone on Earth being handed overnight a connection
to Neal Stephenson's Metaverse *and* the tools with
which to add to it). If most people quit their jobs, cities
everywhere could end up looking like "Burning Man"-style anarchy.
This might all take months or just weeks.

On the human timescale, I think the "haywire period" would
end with a bang, namely with whatever the first fast minds of
"approximately human level" do. I favor the view that they
would intervene along the lines of David Pearce's hedonistic
imperative, while building a "stairway to heaven" whereby
slower entities can take steps into a transhuman existence;
meanwhile, their own accelerated evolution would have made
them worthy of the name Powers, and everything that really
interested them would be taking place in a region reserved
for their own, possibly eternal, future evolution. At least,
all this happens if they don't make a Big Mistake first.

I feel on much shakier ground speculating about this future
of accelerated superintelligences, than I do when spinning a
nano scenario.

Some URLs:

http://www.foresight.org/Conferences/MNT05/Abstracts/index.html
Abstracts for a Foresight Institute conference next month,
focusing on fullerenes

http://nano.xerox.com/nanotech/hydroCarbonMetabolism.html
Merkle's "hydrocarbon assembler"

http://www.zyvex.com - Zyvex

http://nanocomputer.org - The "Nanocomputer Dream Team"

-mitch
http://www.thehub.com.au/~mitch

PS Remember to check out the Mediator! http://crit.org