Re: Freitas on Gray Goo

From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Mon May 15 2000 - 07:30:06 MDT


On Sun, 14 May 2000 hal@finney.org wrote:

> >
> > With sufficient monitoring capacity we would know the *day* anyone
> > released badbots on the planet.
>
> With sufficient monitoring capacity, any information gathering is possible
> in principle. So this last statement is too broad to be very useful.

Ok, I was primarily using space-based observers as lookouts for
accidental releases or mutations. These are primarily point source
so orbiting satellites viewing locations once every few hours to
few days would see the development in time to allow deployment
(or even active defense if you allow large space-based lasers).

>
> But if we are talking about early or intermediate nanotech, and our
> monitoring is restricted to detecting temperature anomalies, I don't
> think this is effective against some of Robert Freitas' scenarios.

No, as I suggested to Robert, but he chose not to include in the
paper you can also to high throughput air & water monitoring using
a distributed network of briefcase sized distribted mass-spec machines.
These would be placed around the globe and linked to satellites (similar
to the seismometer networks currently in use to monitor nuclear
tests). Nanobot activities would distort not only local
temperatures but local atmospheric or water concentrations as well.
Because they are widely placed and could be interlinked with an
understanding of insolation, wind patterns, etc. they should be
able to detect changes in thermal activities well below the 4 degree
threshold Robert uses.

The nanobots would have to be designed *very* carefully
to avoid tipping their hand that they are converting the biomass
into nanomass. (Likely byproducts of that process include things
like N2, NO, NO2, etc.). So replicants would have to keep the
production of these materials below natural concentration variations.

It is useful to note that we are rapidly moving in the direction of
having such monitoring capabilities in place for the purposes of
tracking global warming, extraterratorial pollution, etc.

>
> Assume that the goo seeds are spread widely before beginning their
> growth phase. Robert F. suggests dispersal by airplane. Perhaps it
> would work to spread rock-coated nanites in the form of dust released
> into the atmosphere from a few locations over a period of many months.
> We now have potentially billions or even trillions of nucleation sites.

Ok, so here you are talking about an intentionally designed terrorist
weapon using "advanced" nanotech.

>
> Each of these begins to grow simultaneously, not caring about local
> thermal pollution, but limited by some constraining temperature, such
> as the boiling point of water. (We cook our food before eating it;
> perhaps the goo will find it more tender that way as well.)

I dealt with this as well in my comments to Robert. You simply *do not*
have the resources around the globe (energy or matter) to have uniform
growth rates. Not enough biomass in deserts, arctic regions or under
the ocean. Probably not enough energy in arctic regions or the ocean
as well.

>
> According to Robert F.'s figures, this will not cause a noticeable (4
> degree) global increase in temperatures until something like 0.0001%
> of total biomass has been converted. Due to the wide dispersal,
> localized hot spots may not be detectable either. At this point the
> global temperature will rapidly increase to 100 degrees C.

No, only in regions where the energy and material resources are
sufficiently dense will they be able to grow at the most rapid rates.
There are huge regions of the planet where they will have a difficult
time replicating at all. So you get local hot spots that can be
detected and eradicated from space or using ground based weapons.
The standard precaution against this would be to have defense systems
in places where biomass consuming nanobots cannot grow.

If there is a plausible defense installed, then the construction of
such nasty machines and their distribution would be rather pointless.

> At this point the growth rate ramps down and the remaining conversion
> takes three weeks so we don't burn our food. Of course by this time
> the temperature has killed all life anyway.

Yes, if your goal is the destruction of life on the planet then this
would be a good way to accomplish it except for the reasons I've cited
above. However, the probability that individuals who are that irrational
would have the resources to develop the advanced nanotech and delivery
systems required to produce the scenario you describe seems very low
to me. Their resources would certainly be far less than the resources
of the groups developing the monitoring and defense systems.

> I don't see how particle beams and similar technologies can deal with
> an attack of this sort. The goo sites are too numerous and widespread,
> and the time from thermal detection to total destruction of the biosphere
> is too short.

You do standard fire outbreak containment tactics. You dump a line of
radioactive waste if they are coming across the ground. You go underground
or underwater if they are airborne. As a last resort you simply nuke em.
With the type of advanced nanotech you are discussing we can rebuild
the planet in a matter of days, so if you have to level it to start
over with a more rational group of people, so be it.

Of course the trick is to be members of the groups in space or in the
bunkers if and when those difficult choices have to be made.

Robert's paper does not deal with what the responses should be to such
outbreaks. It simply says that we can detect them and do something
if we are prepared. Being unprepared is kind of like imagining
medicine without stockpiles of antibiotics. -- "Oh, yes, we know
how to make them, but we didn't want to spend the money to stockpile
them because disease outbreaks happen so infrequently." -- Any
prudent person would recognize this as a bad idea.

Robert



This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:28:37 MST