Re: Methane threatens to bake humanity like Turkeys in an Oven

On 4 Feb, 17:17, "George Dishman" <geo...@xxxxxxxxxxxxxxxxx> wrote:
"Ian Parker" <ianpark...@xxxxxxxxx> wrote in message

Ion drive to lift from the Moon? Not even close. From
a small asteroid with microgravity perhaps but you then
still have to get it to L1. In fact what ou would do if
the manufacturing technology was available would be to
land a mass driver on the asteroid, push it to the right
location, then start processing at that point. You don't
want to be moving a vast thin plate around and coping
with the tidal forces as it passes the Earth.

Lift from an Asteroid. Of course you can't have ion drive from the
Moon. As far as manoevering a thin plate is concerned. In point of
fact that would be easier in space than a solid mass. After
manufacture it could be used to concentrate sunlight and fuel the ion
drive.

As far as tidal forces are concerned. These are taken care of in your
gravitational calculation. All spacecraft to the outer solar system
have used slingshots. We are well used to performing the calculations
needed to traverse complex gravitational fields.
What I propose is similar to George Bush's return to the Moon -
only with one difference - No astranauts. The moror skills of an
astronaut can be replicated completely by machines.

We are nowhere close to that level of technology. The
rovers on Mars are the best. If there was an astronaut
there, he could blow the dust ou of the failed motors
and get them working again. The rovers can't even do
that and if they could it would still take a human
operator to control it remotely. A computer could
easily detect the motor had failed but wouldn't be
able to work out why or how to repair it and again we
are many years away from anything like that capability.

How much is the ISS and NASAs infatuations costing? A far lower sum
would bring capabilities up. You are right though to focus on the low
level of achievment of NASA. In fact games distributors are doing
better than NASA. If we break the problem down into its logical
components we have at its bare bones a problem in dynamics. Now
dynamical systems are inherently predictable. They consists of
matrices and transfer functions. CAD/CAM systems already do good
simulations. Robots are now starting to be developed using Javalink. I
will again point to the basic infatuation and failure of NASA.
Progress is being made by CAD/CAM manufacturers, games (quite
surprisingly) Sonic the hedgehog stands a far better chance of
repairing Hubble than ever NASA has and of couse general manufacturing
industry.

As far as working out why a component had failed. This is quite
simple. How would we do it? Well we would probably take a circiut
diagram, put probes on the various points. We would check the
continuity of circuits in (for example) the coils before coming to the
conclusion that the armature had burned out. If we had a new part we
would then repair the motor. If we did not have the part a robot could
not repair it, but then neither could an astronaut. In fact the
example you give is a rather poor one because car firms like Ford
ALREADY have a plug into which you can place a probe and a computer
then dianoses the fault. It is in fact an integral part of an engine
management system. My car has its ABS controlled from EMS and
increasingly a variety of functions are going onto EMS. I could sttick
a probe in drive the car and proform an emergency stop while going
fast round a corner on a wet road. The system will tell me what is
happenning, and any faults with ABS. Expert systems are in fact used
in fault diagnosis including medicine - which is really just another
form of fault diagnosis and repar. The barrier is not the ability of
the system to diagnose it is patient acceptability and the fact that
diaglogue has to be done via NL (elastic station The season of spring
= La estacion de ressorte). In a car we do not use NL and a probe on
EMS is a lot simpler.

How far should Ford take it? It depends on whether they are paying
their mechanics $180 million a week
OK one might say
can you automate everything.

You might, the rest of us have a much better grasp on
reality.

We are not talking here about emotion or any real deep thought
process, we are simply talking about analyzable dynamical systems. In
fact a task can be viewd as the solution of equations something that
has been done for yonks.

Out of what? The closest we have got so far is to assemble
blocks that have already been constructed:

http://www.news.cornell.edu/stories/May05/selfrep.ws.html

That is vastly different from sending a machine to an
asteroid. The simplest chips in those blocks took billion
dollar fabs to manufacture and that is after the silicon
and dopants have been mined, refined and grown into crystals.


No it isn't! Point is that an asembly from pre prepared components can
be viewed as a dynamical task. The basic Pons Assinorum is the ability
to perform a dynamical task. Producing the robot itself (not from pre
prepared components) can be broken down into a series of dynamical
tasks. If our task is in CAD/CAM we have a desciption of it which a
robot can understand. If we can describe an astronaut's task on the
Moon (or better still an asteroid) and can break down all the
processes into simple steps which can be described dynamically we can
construct a Von Neumann swarm. We should really talk about a swarm not
a machine if we are speaking strictly. We can in fact construct a
definition of a VN swarm along the following lines. We have a number
of processes each with inputs and outputs. The inputs could be
assemblies like a flatpack or they could be sunlight and asteroid
rock. Outputs for some processes are inputs for other processes. We
take the sum of the ensemble. When the only inputs, which are not
matchable by outputs is sunlight and rock we will have achieved the
goal. In fact if we have processes floating around in a CAD/CAM system
we need an algorithm (possibly a GA) to get the fittest solution.
Those who are interested in the origin of life might care to ponder
how a similar sort of algorithm might have operated in ponds some 3.8
billion years ago.

Can we go the whole way in one go? I think that as a first stage
complex chips should be blasted from Earth in rockets.

We will achieve what you are considering eventually using AFM

http://www.almaden.ibm.com/vis/stm/atomo.html

http://research.chem.ucr.edu/groups/bartels/

but we are _many_ decades away from using it for macroscopic
objects.

The NASA establishment
seems to believe in something like "Intelligent Design" (As does the
boss) Astraunauts seem to be enddowed with the kind of "vital force"
postualated by (some) 19th century chemists. My contention is that
anything an astranaut can do a machine can do - probably better.

Go ahead then, see if you can get farther than Cornell.
Or just publish the drawings for a self-contained machine
nanotech machine that can make a copy of itself out of a
lump of raw asteroid using AFM without any preprocessing.

I am endeavoring to show how to do it. I am in fact a retired
scientist and my aim is to interest other people in worthwhile
projects. Nanotech is frequently talked about. The best route to VN is
via CAD/CAM. Nanotech is problematic when doing anything other than
reproducing itself. CAD/CAM will make anything compliant.

Stanford and Cornell are doing a great job - don't get me wrong. My
only criticism is that Andrew Ng to take an eample is also qworking on
linguistics - writing some great papers by all account. However we
would move faster with a TOTALLY DEDICATED effort. I think people like
IKEA and B&Q should sponsor students. If I were a university teacher I
would feel the need to teach a balanced AI course. To be a teacher a
Spanish, just as every teacher is "a teacher of English".

As far as total amounts are concerned. This has in fact already been
discussed by other people (see sci.space.policy) 1/10 of the area of
the Earth's circumferance has been discussed. In the case of VN the
weight is largely irrelevant. A non VN Earth transportable solution
with mirrors weighing 55kg/km^2 has been proposed. Rigidity is not in
fact a probem. If you did want a rigid structure you would spin it.

Nope. If you spin it face on between the Earth and Sun
today, it is edge on in three months :-( It will spin
once a year on an axis perpendicular to the ecliptic.

You are right. However as I have said a non rigid solution would be
what you would probably go for.

Think what the forces would be if a CME hit one side of
the shield. Has anyone calculated that or the strength
needed to avoid destruction? Your are talking about an
emormous mass to make this survivable.

NO - we can still think about spin but now in a gedanken sense. The
oder of magnitude of forces at L1 will be > 5km/h. This is not an
enormous force. I talk in terms of speed since the strength required
has the dimensions of speed. I fact it need not be strong at all.


That is the whole point about methane. This will not work.

The point is to reduce the heating below the threshold
for the positive feedback in the methane release to
overcome the natural negative feedback in the rest of
the biosphere.

Fossil sources will run out eventually so it's a bullet
we will have to bite sooner or later. Of course, while
it is a logical solution, it is as feasible politically
as your sunshade is economically.

It is economic. Once a seed is developed it is fire and forget (not
quite you still need to control it).

Sure, but we won't have that sort of technology until late
in the century at best - too late for sure.

How come we will be getting strong AI according to Blair's think tank?


- Ian Parker

.

Relevant Pages

  • Re: Stories Id Like To See Written
    ... > It might be interesting to glance over "The Dynamics Of The Asteroid" ... > to blow up a planet. ... > that perhaps Prof. Moriarty was writing about how asteroids could ...
    (rec.arts.sf.written)
  • Re: Asteroid near miss
    ... seriously whacked by a million tonne asteroid that our crack NASA ... as to its next encounter with Earth or Selene. ... unknown mass still has our name on it. ...
    (sci.astro.amateur)
  • Re: Am 4. Mai 2102: Nasa - Weltuntergang in 96 Jahren!
    ... Die Nasa geht keineswegs davon aus, dass am 4. Mai 2102 die ... Welt untergeht. ... Ein riesiger Asteroid ist unterwegs zu unserem ... "Bild" schreibt: ...
    (de.talk.tagesgeschehen)
  • Re: NASA asteroid scientists versus 13-year-old.
    ... NASA Statement on Student Asteroid Calculations ... who claims the Apophis impact probability is ... collision with a satellite is exceedingly remote. ...
    (sci.space.history)
  • Re: In the bulls-eye
    ... >> And yet asteroids hit the earth. ... Nasa assumed the following: ... > orbital hit with a precision less than the asteroid. ... comet like and harder to do if 2004 MN4 is iron-nickel but still ...
    (sci.physics)