Re: poly: The singleton hypothesis

From: Nick Bostrom <>
Date: Fri May 29 1998 - 18:14:58 PDT

Anders Sandberg wrote:

> For an interesting example of how this might work, look at virtual
> reality systems. The military VR systems are *behind* many of the
> civilian commercial systems since they tend to be developed in fairly
> closed environments. This is actually even true for academic VR
> research - while the concepts are great and can explore many new
> directions, they are clunky and often behind that is done at
> commercial VR companies.

That's an interesting data point. I don't know enough about VR to

> > The question is whether, when the breakthrough
> > is approaching, the military labs can push a head a little bit, at
> > least regarding the military applications. This seems quite likely.
> > (1) The civilians might not have much interest in the military
> > applications;
> Actually, civilians might be very interesting in countering military
> applications, like developing protections rather than weapons.

Sure civilians would be interested in good defenses, but as with
other military systems they typically leave it to the military to
develop and manage them. That's what's the military is for, after
all. The military may of course commission R&D from private

Note also that if civilians try to develop active shields and the
military try do develop black goo, the military probably has the
easier task.

> > (2) They might even be banned from doing research on
> > these aspects
> Hard to do in a global research community, especially with *defensive*
> research.

It will be hard to sharply separate defensive from
potentially offensive research. The enabling technologies are the
same in both cases. And rather than allowing anybody to play with
potentially destructive nanotech in their garages (to allow that
before we have developed active shields would be absolute madness)
govenments might choose to ban private nanotech research altoghether.

> > (3) An organization like the US military could muster
> > enormous finansial resources if it thought US security critically
> > depended on coming first to nanotech; it could buy up most
> > researchers doing open research.
> Could it? I think this is the mythical man-month idea, if you need
> something done tomorrow, hire enough people to do it. Mongolian horde
> tactics doesn't work well in research and development - people poured
> money on SDI, and didn't get much. And buying up most resarchers
> doesn't seem feasible unless there is a very small number (in
> addition, other groups would be highly motivated to buy up the others
> or help them remain independent if they perceived the US military as a
> threat/problem).

Yes, the militaries of other countries...

Anyway, I want to emphasize that the singleton thesis is totally
independent of the "the military will get there first"-view. A
collaborative effort of the world's civilian research labs, regulated
by democratic organizations, would have a chance of beating the
military, and this seems like something we should strive for. But
there would still be a singleton, only now it would be based on
a democratic organization (or coalition of organizations,
governments and NGOs). What would make it a singleton is that it
would prevent other groups (or sub-groups) from
independently developing threatning technologies. It could offer
non-members to join, for ethical reasons; it could abstain from
harming societies who don't join, also for ethical reasons. But it
would be rational for the give up its dominance.

> My impression is that the singleton hypothesis rests on the assumption
> of dominant technology

The singleton thesis does not itself depend on that assumption,
although most arguments in its favour that I have put forward
on this list do. Even if all technologies developed in very small
steps over a very long period of time, it is still perfectly
conceivable that a singleton will form. Nations could gradually grow
together; nation states may wither away; the UN can be strengthened;
a single religion or political view may become universally accepted.
Threre are many possibilities, some more plausible than others.

Such a singleton might then become stable when it develops the
science and technology to, e.g. choose the genetic traits of new
citizens; improvend surveilance technologies; implanted electrodes
for motivational control etc. A global dictator with such tools at
hand might be able to create an evil empire that could never be
overthrown. On the other hand, a free, democratic world order could
also perpetuate itself in this way.

>, i.e. some single technology or method that
> gives you total power, especially in the respect of preventing anyyone
> else from developing it. There is no evidence that such a technology
> is possible, or likely to be developed.

While it's difficult to 100% certain, I think it definitely looks as
if the nanotech+superintelligence breakthrough, as outlined by
Drexler, will be such a dominating technology. It will happen
quickly, and the force who gets it first will have total power.

Nicholas Bostrom
Department of Philosophy, Logic and Scientific Method
London School of Economics
Received on Sat May 30 00:22:43 1998

This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:30 PST