Re: The Singularity and Nanotechnology

Dr. Rich Artym (
Sun, 29 Sep 1996 08:55:28 +0100

In message <>,
Crosby_M writes:

> I think the majority would agree (or at least like to believe) that
> strong nanotech capabilities will _eventually_ be achieved. But I can
> recall at least two others besides myself who posted concerns about your
> vision of how we can get there from here and what shape it might
> ultimately take.
> As an engineer, and a software one at that, I would think you'd
> recognize the need for calibration and specification that Lyle suggests,
> as well as the practicality of incremental development. Too many
> 'grand-scheme' or 'master-plan' software development efforts have
> crashed & burned.

OK, this is a reasonable response. Since you accept that MNT will
_eventually_ deliver, we can at least argue the whens and whats.

The reason why I am a front-line optimist on the MNT rate-of-change front
(but have large concerns about whether we'll survive, which follows from
the rate-of-change argument) is that, unlike the state of the world when
the transistor was developed, today we have engineering and computing
facilities that will collapse the time required to a fraction of the 50
years it took us to develop million-transistor integrated circuits.

So many technologies are converging to deliver the required engineering
capability that it's hard to keep up. Every time we look around there's
been yet another DNA sequenced or protein created, semiconductor plants
are driving back the barriers of the smallest features they can create
mercilessly, fuzzy neural companies are increasing the cleverness of
their gizmos all the time, and every man, woman and their dog now has a
top engineering workstation on their desk with the capability to create
any design they wish. So many thousands of software teams are delivering
full working packages to the public domain weekly as seen in the many
systems/languages announce newsgroups that the capability of the man in
the street is undergoing an explosive increase that exceeds even that
of many a corporate researcher bound by company standards. We're well
onto the lower slopes of that good ol' Singularity.

That simply cannot fail to have an impact on the rate of progress of
MNT development, it seems to me. Many different approaches are being
taken by MNT researchers so it's not a case of all the eggs being in
one basket, and the software design systems can adapt to whichever
hardware route proves viable. Furthermore, as many people have said,
whichever hardware route succeeds in delivering an elementary building
block, from that point on we can start to bootstrap ourselves out of
the (probably soft) embryonic technology into other forms. Of course
there are numerous unknowns in this area, everyone accepts that, but
the incredible synergy around right now at such an early stage makes
me think that MNT will happen sooner rather than later. As others
have said, why *should* it take 50 years, what are people going to
do the other 30? :-)

Much of the time estimate relies on "feeling" based on seeing what's
happening around us, aided with a sense of what engineering research
is about. Fair enough, "feeling" differs between people, so others
may be working to different estimates, I accept that. Given the 20-
year figure though (and that's not just mine, I'm merely agreeing with
the Drexlerian 10-15-20 year scenario), that represents the blink of
an eyelid in social and economic terms, which is why I insist that the
development of MNT harbours not an evolution of social and economic
systems but a cataclysmic revolution.

Mankind is capable of accepting anything (I believe) as evolutionary
given time, but I do not think that he's going to have the time to
make a smooth transition. If we're talking about 20 years then there
is no chance at all of evolution in economics --- shucks, most board
members in the corporate world haven't yet come to terms with the
Internet and those nerdy computer things, and they *don't* read SF. (:-)

The status quo is so strongly rooted in a historical legacy of slowly
changing systems with the institutions of the status quo in control
that there is no prospect of it changing gracefully to meet MNT as
a challange. Instead, I expect MNT to arise out of a combination of
small engineering concerns, academic research departments, privateers
and the massed ranks of the public domain, and then to suddenly and
very dramatically explode on an unprepared world. My *hope* is that
it *MERELY* causes the entire economic system to collapse for financial
reasons; my fear is that it will be far worse than that, and that even
if the biosphere doesn't get converted into goo, many inimical forms
will nevertheless be released, and we may not survive it.

[By the way, the reasons for the mentioned "collapse" I have already
explained several times, so I won't repeat them here. No economics
detractor that *does* accept MNT (Lyle doesn't) has yet come back with
any argument at all explaining how current-day economics can survive
without a dependent customer base to feed money through the system;
the message about means of production appears not to have got through.]

> IMHO, the scenarios you've suggested would be feasible only for some
> stripped-down, 'virtual nanotechnology' simulated in an Artificial
> Reality, OR in a world where nanites were already in control of
> everything - "an organic/silico blob" / "a teaming, feasting soup" as
> some posts on older threads envisioned. That day may come; but, if
> you're really "more interested in *making* it happen", you should be
> working on a practical transition plan.

It's the practical transition plans that are doomed to failure: this
MNT stuff has no prospect of being developed in a controlled manner,
as it's kitchen-table technology in principle. Just because we may not
like "a teaming, feasting soup" doesn't mean that we won't get one.

I don't get the "virtual nanotech" reference. The elements needed for
MNT to develop along well-explained paths are being created now, today,
in the real world. Anyone that says it's only possible in virtual
settings have their heads rammed deep, deep into the sand.


###########  Dr. Rich Artym  ================  PGP public key available
# galacta #  Email   :
# ->demon #  Web     :
# ->ampr  #  AMPR    : rich@g7exm[.uk] BBS:GB7MSW
# ->NTS   #  Fun     : Unix, X, TCP/IP, kernel, O-O, C++, SoftEng, Nano
###########  More fun: Regional IP Coordinator Hertfordshire + N.London