Human spaceflight and AI

h (Rand Simberg) wrote in message ...

snip

Yes, it says that, contrary to your amusing and senseless notion, we
won't commit mass suicide just because AI is developed.

On the other hand, it becomes economic to come up with ways to reduce
the human population humanely via control over reproduction. Over a
period of ten to twenty generations we should be able to reduce the
population to a level which still maintains genetic and cultural
diversity. I think ten million is a reasonable number with minimal
impact on Earth's ecological systems and if control over reproduction
is lost (say in the unlikely event that the human population revolts,
we have perhaps 8 to 10 doublings before the carrying capacity of
Earth is reached. Plenty of time to restore control.


Karl Hallowell

"Karl Hallowell" wrote in message
om...
h (Rand Simberg) wrote in message
...

snip

Yes, it says that, contrary to your amusing and senseless notion, we
won't commit mass suicide just because AI is developed.

On the other hand, it becomes economic to come up with ways to reduce
the human population humanely via control over reproduction. Over a
period of ten to twenty generations we should be able to reduce the
population to a level which still maintains genetic and cultural
diversity. I think ten million is a reasonable number with minimal
impact on Earth's ecological systems and if control over reproduction
is lost (say in the unlikely event that the human population revolts,

Unlikely?!!

we have perhaps 8 to 10 doublings before the carrying capacity of
Earth is reached. Plenty of time to restore control.

And precisely what entity will "restore control"?



Karl Hallowell

Pascal Bourguignon wrote in message .. .

Please, follow the thread! Here is the original post:


From: (Alexander Sheppard)
Newsgroups: sci.space.policy
Subject: Human spaceflight and AI
Date: 6 Feb 2004 19:55:19 -0800
Organization: http://groups.google.com
Message-ID:

I was just thinking, there may be a rather simple point that will make
human spaceflight more or less irrelevant. Suppose at some point,
quite possibly in just a couple of decades, we develop computers that
have the intelligence of human being. Then why not send AI there
instead? There would simply be no economic reason, except tourism, for
sending humans into space. An artificial entity would could be made
infinitely more suited for any task, especially with the help of
advanced nanotechnology.

There would simply be _no_ _economic_ _reason_ why we should still
live. After all, these AI would be much more efficient than us at
living! (For example, they could feed directly from silicium and sun
instead of relying on complex biospheres, and what's more, they could
change their body at will to adapt the conditions. They could even
teletransport: sending their minds to bodies on other planets thru
electromagnetic waves).

I think we wouldn't all want to commit suicide. But perhaps enough
people would like to live in a VR world and explore the Universe, and
live perfect lives. Eventually, to avoid disruption to the matrix, all
humans would have to participate.

(Karl Hallowell) wrote in message . com...
h (Rand Simberg) wrote in message ...

snip

Yes, it says that, contrary to your amusing and senseless notion, we
won't commit mass suicide just because AI is developed.

On the other hand, it becomes economic to come up with ways to reduce
the human population humanely via control over reproduction. Over a
period of ten to twenty generations we should be able to reduce the
population to a level which still maintains genetic and cultural
diversity. I think ten million is a reasonable number...

Practically the only reason population growth is a concern now
is that the standard of living the world's population aspire to
is what is had in places like the U.S. and current standard
technology requires that standard of living to make heavy demands
on the environment such as by coal burning electricity producing
plants and smoke spewing automobiles.

If we wanted to we could have a global or near global power
grid fed by geothermal and other clean types of power producers.

[...] with minimal
impact on Earth's ecological systems and if control over reproduction
is lost (say in the unlikely event that the human population revolts,
we have perhaps 8 to 10 doublings before the carrying capacity of
Earth is reached. Plenty of time to restore control.

Oh, you were just joking.

-McDaniel

Michael Gallagher wrote in
:

On 7 Feb 2004 11:05:41 -0800, (Alexander
Sheppard) wrote:

Our robotic machinery is
sometimes just not up to some things. I recall on the space shuttle
there have been many cases like that. So the astronauts go out and fix
it ....

IIRC, on one Shuttle mission, while the astronauts were asleep,
Houston sent up commands to stop a tumble the orbiter was in, but it
started rolling as well as tumbling. They tried to null all that out,
and instead of rolling and tumbling, it was rolling, tumbling, and
yawing.

Hmm, not quite. The flight was STS-32. The orbiter was initially stable,
not tumbling. MCC was not sending commands to stop a tumble; they were
performing a routine state vector update (the state vector tells the nav
software the orbiter's position and velocity). The update was botched due
to human error by a flight controller. The resulting state vector was
binary garbage, causing the orbiter's nav software to think it had suddenly
teleported very, very far from Earth. Since the orbiter was holding an
Earth-pointing attitude at the time, that's what kicked off the tumble.

So they woke up the mission commander who went up to the flight deck,
probably 'blessed' Houston on seeing the Earth doing funny cart wheels
outside the window, nulled everything out in under five minutes, and
went back to bed.

That part was right. The commander was Dan Brandenstein, by the way.

--
JRF

Reply-to address spam-proofed - to reply by E-mail,
check "Organization" (I am not assimilated) and
think one step ahead of IBM.

(Hobbs aka McDaniel) wrote in message . com...
(Karl Hallowell) wrote in message . com...
h (Rand Simberg) wrote in message ...

snip

Yes, it says that, contrary to your amusing and senseless notion, we
won't commit mass suicide just because AI is developed.

On the other hand, it becomes economic to come up with ways to reduce
the human population humanely via control over reproduction. Over a
period of ten to twenty generations we should be able to reduce the
population to a level which still maintains genetic and cultural
diversity. I think ten million is a reasonable number...

Practically the only reason population growth is a concern now
is that the standard of living the world's population aspire to
is what is had in places like the U.S. and current standard
technology requires that standard of living to make heavy demands
on the environment such as by coal burning electricity producing
plants and smoke spewing automobiles.

If we wanted to we could have a global or near global power
grid fed by geothermal and other clean types of power producers.

You don't have to try as hard if there's much less people around.
Besides this branch of the thread got started because as a
hypothetical situation. That is, due to superduper AIs or other wonder
technology, there is literally no economic reason to buy a good or
service from a company which employs humans and no reason to sell to
poor humans. In other words, each human is a net cost to society
(which consists of AIs, etc as well).

We can cram a lot more people on the planet or in space, but if humans
are a net cost to society, then even ethical societies will figure
ways to reduce the problem. The obvious way is through population
reduction. Alternately, they might convert humans to a less costly
form. Eg, the practice of "uploading" human conciousness into
wondertech.

[...] with minimal
impact on Earth's ecological systems and if control over reproduction
is lost (say in the unlikely event that the human population revolts,
we have perhaps 8 to 10 doublings before the carrying capacity of
Earth is reached. Plenty of time to restore control.

Oh, you were just joking.

Sounded reasonable to me. But sometimes the funniest jokes weren't
jokes... were they?


Karl Hallowell

(Karl Hallowell) wrote in message . com...
(Hobbs aka McDaniel) wrote in message . com...
(Karl Hallowell) wrote in message . com...
h (Rand Simberg) wrote in message ...

snip

Yes, it says that, contrary to your amusing and senseless notion, we
won't commit mass suicide just because AI is developed.

On the other hand, it becomes economic to come up with ways to reduce
the human population humanely via control over reproduction. Over a
period of ten to twenty generations we should be able to reduce the
population to a level which still maintains genetic and cultural
diversity. I think ten million is a reasonable number...

Practically the only reason population growth is a concern now
is that the standard of living the world's population aspire to
is what is had in places like the U.S. and current standard
technology requires that standard of living to make heavy demands
on the environment such as by coal burning electricity producing
plants and smoke spewing automobiles.

If we wanted to we could have a global or near global power
grid fed by geothermal and other clean types of power producers.

You don't have to try as hard if there's much less people around.

Who says we're trying 'hard' now? When I turn on the light switch
at my house I don't have to pour fuel into a generator, chop lumber
or turn a hand crank on a DC power generator. To me personally there
is probably more hardship in trying 'less' than there is in trying
'hard'. I'm sufficently isolated from the production side of
electrical power due to the multi-layered distribution infrastructure
for that to be the case. My only hardship is to mail a check or
setup a direct bank account withdrawl for the power distributor.

Oh, you meant 'we' as in the species? I don't think it's in my
interest to do what's best for everybody else... although I'm
willing to do things that are necessary as opposed to simply
pie in the sky what-if experiments.

Besides this branch of the thread got started because as a
hypothetical situation. That is, due to superduper AIs or other wonder
technology, there is literally no economic reason to buy a good or
service from a company which employs humans and no reason to sell to
poor humans. In other words, each human is a net cost to society
(which consists of AIs, etc as well).

If the moon had a nose we could put really big sunglasses on it too.

We can cram a lot more people on the planet or in space, but if humans
are a net cost to society, then even ethical societies will figure
ways to reduce the problem. The obvious way is through population
reduction. Alternately, they might convert humans to a less costly
form. Eg, the practice of "uploading" human conciousness into
wondertech.

Of course the proponets of population reduction will probably be
the same bureaucrats who spend billions supposedly helping the
poor while their own pension funds and office budgets baloon.

[...] with minimal
impact on Earth's ecological systems and if control over reproduction
is lost (say in the unlikely event that the human population revolts,
we have perhaps 8 to 10 doublings before the carrying capacity of
Earth is reached. Plenty of time to restore control.

Oh, you were just joking.

Sounded reasonable to me. But sometimes the funniest jokes weren't
jokes... were they?

Well I find it laughable to reduce the human population to 10k people
as an earlier poster suggested. He pulled that number out of his
rear. I can think of a billion reasons why that'd be a bad idea.
For instance you can't simply put the whole human population in the
same city because it would be vulnerable to loss by local natural
disasters. If you spread them out too far it could have unforseen
consequences on the culture (not that reducing earth's population
to 10k wouldn't!) What percentage of those 10k people are going
to make good scientist? Artists? Medical doctors? And so on.
What happens when too many of the key members die sooner than
expected? Is there the genetic diversity to weather an epidemic
from an as yet unknown virus?

Oh.. I forget. AI - that magic elixar -- will solve any and
everything.

-McDaniel

(Alexander Sheppard) wrote in message . com...
I was just thinking, there may be a rather simple point that will make
human spaceflight more or less irrelevant. Suppose at some point,
quite possibly in just a couple of decades, we develop computers that
have the intelligence of human being. Then why not send AI there
instead? There would simply be no economic reason, except tourism, for
sending humans into space. An artificial entity would could be made
infinitely more suited for any task, especially with the help of
advanced nanotechnology.

Computers in the 1950s were things to wonder at. At that time IBM
said that the whole world market for computers might be a dozen
machines maybe a market for one or two really powerful ones (16 KB
electronic memory!)

Of course, things have changed dramatically in computing over the past
55 years. That's because Noyce and Moore saw that by reducing the
cost of circuitry the demand for circuitry increases to more than
cover the R&D to make that reduction in price possible. As a result,
computing has grown exponentially over the past 55 years at a very
rapid rate, and is likely to continue growing for the next 45 years,
given the size of atoms and the energetics of information processing.
Futurists like Moravec and Kurzweil predict that by 2030 computing
systems will be less costly and more capable than human beings in any
economically useful task. By 2050 we will have entered an era of
superhuman computing. Mere humans cannot predict -without the
assistance of superhuman computers- what that might mean in human
affairs.

Now, the reason I mention this development arc in computers is it
informs us what has to be in place for radical transformations to
occur in the present situation. These are;

(1) A reliable measure of performance - a metric, (cost/computation)
(2) A means to improve that metric in a cost-effective way
(increases
in market size pay for R&D)

The first question therefore wrt space travel becomes; What is a
reliable measure of performance? The answer is - COST/MOMENTUM, or
dollars per kg-km/sec. Reduce this cost and the cost of achieving
things in space are lower.

The second question now becomes; Can we improve that metric in a
cost-effective way?

Well, here we have some arguments. First, this metric has not been
improving over the years. Some would have us believe this is a
fundamental rule of nature. But this is plainly not the case. Our
failure to improve the momentum metric has many causes, all of which
can be changed. THese include;

(1) Cost plus contracting - wherein higher costs are rewarded with
higher profits.

(2) Missile proliferation concerns - wherein the risk of misuing
missiles by people we don't like (and who don't like us) is considered
greater than the benefits of widespread use of rocket technology. A
very inexpensive rocket would basically spell the end to missile
proliferation controls. This is a bad thing in many intellignece
officer's minds.

(3) Upsetting the defense/high-ground apple cart - Expanding human
presence beyond the surface of the Earth means that current defense
and intelligence gathering hardware focused on the surface of the
Earth become ineffective for an important segment of the human race.
A segment that has the capacity to affect Earth's population in
significant ways. This is a bad thing in many military minds.

But, setting these concerns aside, is there a way to make space-travel
pay? Well, that depends on our ability to find markets that pay for
the tremendous costs of space travel, and the tremendous R&D cost of
reducing the cost of space travel.

Do such markets exist?

Certainly.

At present we benefit greatly from global communications networks.
These networks would not be possible without communications
satellites. US News & World Reports indicated in 1996 that the annual
amount of money generated by communication hardware in space was $83
billion. This more than paid for the $12 billion a year the world
spends to launch communication satellites. Given the limited way in
which we use comsats and the advances in communication technology, we
could expand this role of comsats greatly well beyond the hundreds of
billions of dollars per year. Communications in space proceed as
follows;

(1) One to one - telstar
(2) One to many - DirecTV
(3) Many to many - Iridium/Teledesic

Teledesic was put on hold after the failure of Iridium. But, the
direction of success is clear. Greater flexibility of communication
and broader bandwidth. Ultimately we will have a wireless broadband
internet connection for everyone - and multiple such connections for
wealthier folks - everywhere. With this communications link not only
will we be capable of communicating unprecedented amounts of
information, but we will also be capable of routinely using
telerobotics for work and pleasure. With telerobotics space
communications could rise into the trillions of dollars per year -
more than enough to prime the pump of rocket development.

Beyond the communications revolution that's possible with space based
assets, there is a potential for a revolution in the way we make and
use power. P. Glaser in 1968 proposed sending up a new kind of
satellite. A power satellite. This powersat would collect sunlight
in space and generate electricity. It would then beam this
electricity back to Earth for use there. Since there is no night in
space, electricity could be generated 24/7 without any storage needed.
With advanced interferometry and laser technologies it will be
possible to beam either laser or microwave energies very reliably and
safely - despite high beam energies - to anywhere power is needed on
Earth. While current energy expenditures are limited to around $650
billion per year now, energy use is a fundamental determinant in
economic growth. So, if power can be delivered more cheaply by
beaming from space than from fuels sent around the planet by cargo
ships, we are setting the stage for an unprecedented spurt in economic
growth. At the same time we're ending our pollution of the air and
water and ground due to burning of fossil fuels. In a world where
everyone is consuming energy at the same rate as Americans total
demand for energy could exceed $5 trillion per year.

As a side benefit of this economic growth its important to note that
when per capita income rises above $10,000 per person per year,
population growth rate falls below replacement level. So, at $30,00
per person per year - average (there will still be poor and rich in
the future) average population growth world wide will be negative.

Low cost heavy lift launchers are a requirement for Earth based
construction and launch of solar power sattellites.

Beyond comsats and powersats there are other possibilities.

Nuclear pulse spacecraft have been proposed since the 1940s. These
spacecraft use a series of sanitized nuclear explosions to drive huge
spacecraft to unprecedented speeds. Spacecraft the size of cities.
Some have suggested using this technology we could have saved the
dinosaurs from extinction by deflecting the asteroid that killed them
all. We could do this again to save ourselves if we had sufficient
warning. This is true.

But more important is our ability to move rich asteroids into stable
orbits around Earth and then using telerobotics and abundant labor on
Earth - mine the rich reserves of material and process it to finished
goods - using solar power - and deliver a rich array of things
anywhere its needed.

This manufacturing capacity - factory satellites - would also be used
to expand our presence in space! We could build billions of large
pressure vessels and create vast tracts of farmland and forests. In
this way we could create food and fiber for distribution to anyone on
Earth in nearly unlimited quantity - by today's standards. At the
same time we could close down mines, refineries, processing plants,
assembly plants, farms and commercial forest operations from the
surface of the Earth. Humanity could live very abundantly in a vast
global nature preserve supplied from space resources.

In such a world the amounts of money earned in space could grow to
over a quadrillion dollars.

At these levels we could consider building large residential tracts on
orbit. Not small capsules, not bungalows stacked atop one another -
but large geographic areas with rich ecologies the size of counties,
owned by individuals. These would be placed in stable orbits by the
billions between two asteroids that are parked as shepard moons.
People would move to a space residence like this because it would be
far larger than anything they could afford on Earth, and products
would cost far less than on Earth.

Ultimately, the bulk of humanity will live on orbit in large space
homes leaving the Earth's surface to the vacation market and those few
who would stay on Earth for emotional reasons.

Here our Earth based models break down. So, its not clear how
economic growth would be served by continued space growth.

But, its clear that we can imagine business models that make good use
of all our rocket technologies that are capable of reducing cost of
momentum - to expand economic opportunities to unbelievable levels and
bring about important changes in the human condition by imaginative
application of space technology.

(Hobbs aka McDaniel) wrote in message . com...

snip

Well I find it laughable to reduce the human population to 10k people
as an earlier poster suggested. He pulled that number out of his
rear. I can think of a billion reasons why that'd be a bad idea.
For instance you can't simply put the whole human population in the
same city because it would be vulnerable to loss by local natural
disasters. If you spread them out too far it could have unforseen
consequences on the culture (not that reducing earth's population
to 10k wouldn't!) What percentage of those 10k people are going
to make good scientist? Artists? Medical doctors? And so on.
What happens when too many of the key members die sooner than
expected? Is there the genetic diversity to weather an epidemic
from an as yet unknown virus?

Oh.. I forget. AI - that magic elixar -- will solve any and
everything.

Thus ends the great AI thought experiment. I lack sufficient
motivation to continue the struggle against increasingly logical and
rational counterarguments.


Karl Hallowell