Methane threatens to bake humanity like Turkeys in an Oven

On 22 Feb, 11:52, "George Dishman" wrote:
This would be true apart from the methane. If you are asking another
question. Do you have a space microwave system? One can say that the
production of hydrogen in the worlds deserts is far more economic. -
True. However the original point was METHANE. Hydrogen in deserts will
do nothing for that.

All my suggestions reduce the amount of fossil carbon
being introduced into the atmosphere which is the key,
if you get the heating from C)2 down, the stored
methane will not get released, or at least the amount
will be reduced. You break the positive feedback loop.
So yes, solar energy produced in the desert _will_
help regardless of whether hyrogen or a power grid is
used to transport it.

True - however the time constants for mathane etc. are very long.
Longer in fact than the timescales for te chnological development.

I agree that we have probably rxhausted the aruments. There is however
one twist. A think tank commissioned by Tony Blair has talked about
STRONG AI (robot rights) before 2050. Strong AI will mean all the
concomitants of weak AI. You cannot have it both ways.

Whether we have AI or not is irrelevant, and yes I have
dsaid this repeatedly and you always ignore it. The
problem is not the software, it is the ability to
manipulate raw materials that is missing. Remember "An
infinite number of Sonic the Hedgehog cannot bend a
paper clip."

Look - I may be prepared to concede some points on timescales, but AI
MUST be the key. Look if you can send a robot with all the
manipulative properties of an astronaut, you MUST be able to have a
robotic colony. If AI does NOT produce a VN swarm then astrnauts MUST
be a wate of time too. I mentioned Sonic in particular because his
software is going to be used by the Pentagon in mititary robots. I
find it interesting too that military technology is no longer the
leadimng technology. Games and animation seem to be.

Being able to do somthing, no matter what it is, is a matter of being
able to predict the consequences of actions. This is why games are
relevant.

2) Ray Kurzweil says that we will have a singularity by 2030-40.

That would already be too late for this particular
problem. Others are longer term though.

I
think he is over optimistic.

Well so do I. What I am going to say now is not relevant to VN
machines but sure is for the singularity. Google tell us that no more
than 600MB of the human genome is devoted to intelligence. I have been
going to Google Translate and translating into a 2L (Spanish). ?
Quieres dormir con fosforo? That was how "match" in the sense of a
dating match was translated. You basically need mathematical theories
not hand waving arguments.

In by dynamic principles I am proposing hard matematics.

Grossly, as is Drexler. Limited AI by then probably but the
singularity needs a lot more breakthroughs.

Indeed
You say "been and gone". We can of course try to live with global
warming. There is yet another twist on hydrogen. If lands are going to
become arid you can use solar energy to desalinate water.
Delaslinatioon, as well as hydrogen production does not have to be
24/7 unlike general power generation.

Power generation doesn't either, there are deserts around
the world and hydrogen fuel cells together with electrolysis
make a compact if inefficient storage mechanism.

If you are desalinating or smelting aluminium you would like direct
generation and the Sun to be shining. This is the most efficient way.
If you are doing this on a large scale this might not matter.


- Ian Parker

"Ian Parker" wrote in message
oups.com...
On 22 Feb, 11:52, "George Dishman" wrote:
This would be true apart from the methane. If you are asking another
question. Do you have a space microwave system? One can say that the
production of hydrogen in the worlds deserts is far more economic. -
True. However the original point was METHANE. Hydrogen in deserts will
do nothing for that.

All my suggestions reduce the amount of fossil carbon
being introduced into the atmosphere which is the key,
if you get the heating from CO2 down, the stored
methane will not get released, or at least the amount
will be reduced. You break the positive feedback loop.
So yes, solar energy produced in the desert _will_
help regardless of whether hyrogen or a power grid is
used to transport it.

True - however the time constants for mathane etc. are very long.
Longer in fact than the timescales for te chnological development.

No, that's the point. CO2 timescales are long - the
heating effect of what is already in the atmosphere
will continue into the next century, but methane is
short lived. nethane burns in air, CO2 doesn't, and
it will react with the oxygen and be destroyed
comparatively rapidly. Go back and read the original
article again.

I agree that we have probably rxhausted the aruments. There is however
one twist. A think tank commissioned by Tony Blair has talked about
STRONG AI (robot rights) before 2050. Strong AI will mean all the
concomitants of weak AI. You cannot have it both ways.

Whether we have AI or not is irrelevant, and yes I have
dsaid this repeatedly and you always ignore it. The
problem is not the software, it is the ability to
manipulate raw materials that is missing. Remember "An
infinite number of Sonic the Hedgehog cannot bend a
paper clip."

Look - I may be prepared to concede some points on timescales, but AI
MUST be the key. Look if you can send a robot with all the
manipulative properties of an astronaut, ...

Again, that's the point, we don't have robots with
anyhting like those manipulative abilities.

.. you MUST be able to have a
robotic colony. If AI does NOT produce a VN swarm then astrnauts MUST
be a wate of time too. I mentioned Sonic in particular because his
software is going to be used by the Pentagon in mititary robots. I
find it interesting too that military technology is no longer the
leadimng technology. Games and animation seem to be.

I picked up on your mention of Sonic as mostly tongue-
in-cheek but it was intended to emphasise the fact that
no amount of intelligent software will get round that
manipulation problem.

Being able to do somthing, no matter what it is, is a matter of being
able to predict the consequences of actions. This is why games are
relevant.

No, being able to do something depends on connecting
with the physical world.

2) Ray Kurzweil says that we will have a singularity by 2030-40.

That would already be too late for this particular
problem. Others are longer term though.

I
think he is over optimistic.

Well so do I.

I agree, but you wrote that.

You say "been and gone". We can of course try to live with global
warming. There is yet another twist on hydrogen. If lands are going to
become arid you can use solar energy to desalinate water.
Delaslinatioon, as well as hydrogen production does not have to be
24/7 unlike general power generation.

Power generation doesn't either, there are deserts around
the world and hydrogen fuel cells together with electrolysis
make a compact if inefficient storage mechanism.

If you are desalinating or smelting aluminium you would like direct
generation and the Sun to be shining. This is the most efficient way.
If you are doing this on a large scale this might not matter.

The point is that there are economically viable processes
that we can use here on Earth which will not only solve the
problem but do it quicker _and_ can make a profit. The cost
of a basic reflective sunshield would be astronomical and
would provide no financial return whatsoever. Not only that
but we don't have the technology to bring the millions of
tonnes needed into Earth orbit let alone make a shield out
of it.

George

On 23 Feb, 15:21, "George Dishman" wrote:

True - however the time constants for mathane etc. are very long.
Longer in fact than the timescales for te chnological development.

No, that's the point. CO2 timescales are long - the
heating effect of what is already in the atmosphere
will continue into the next century, but methane is
short lived. nethane burns in air, CO2 doesn't, and
it will react with the oxygen and be destroyed
comparatively rapidly. Go back and read the original
article again.

The time constant is caused by the fact that there is methane in
permafrost. As this thaws methane gets released. There has been
speculation about the cause of ice ages. The simnple fact may well be
the unstable equilibrium caused by methane. The fact that methane
oxidises means that it has a finite life, but that lifetime (in lower
than ignition quantities) may not be that short.

Look - I may be prepared to concede some points on timescales, but AI
MUST be the key. Look if you can send a robot with all the
manipulative properties of an astronaut, ...

Again, that's the point, we don't have robots with
anyhting like those manipulative abilities.

I will comment on this with general AI comments.
.. you MUST be able to have a
robotic colony. If AI does NOT produce a VN swarm then astrnauts MUST
be a wate of time too. I mentioned Sonic in particular because his
software is going to be used by the Pentagon in mititary robots. I
find it interesting too that military technology is no longer the
leadimng technology. Games and animation seem to be.

I picked up on your mention of Sonic as mostly tongue-
in-cheek but it was intended to emphasise the fact that
no amount of intelligent software will get round that
manipulation problem.

Being able to do somthing, no matter what it is, is a matter of being
able to predict the consequences of actions. This is why games are
relevant.

No, being able to do something depends on connecting
with the physical world.

Let us do a little thinking round this problem Suppose we built a
robot with similar joints to a human and with electric motors or
fluidic circuits to activate them. Personally I like the idea of a
fluidic robot woth one central pump and pistons for limbs. Building
this (without AI) is a fairly trivial extercise. We could provide
pressure sensors for the finger tips to provide a general "feel".
There is nothing here, except of course for the AI element which could
not be built by a reasonable tean of engineering undergraduates.

Of course without a brain our assembly could not do anything useful.
If it had radio transmittor/receivers it could do a range of things
with telepresence but it would not be capable of acting autonomously.
Put in AI and it then becomes an autonomous machine that you could
send up to repair Hubble, and if in space would do everything an
astronaut could do - and more as we know that robots in medicine are
capable of cutting with a precision that no surgeon can manage.

That is why I talk about AI (need only be weak) as being the key to
achieving Von Neumann + other robotic tasks. Of couse the
"understanding" we need is different from "?Quieres dormir con
fosforo?". CAD files of course form a basis for geometrical and
physical understanding. This sentence BTW is quite an embarassment to
singularists as it does imply a certain amount of the "emotion" they
talk about!

Let us think a little bit more deeply about how we might repair
Hubble. We have of course Hubble CAD files. AI would then be able to
tell us the sort of robot we would need for the task. Chances are it
would be far simpler than the full anthropomorhism I have spoken about
above.


If you are desalinating or smelting aluminium you would like direct
generation and the Sun to be shining. This is the most efficient way.
If you are doing this on a large scale this might not matter.

The point is that there are economically viable processes
that we can use here on Earth which will not only solve the
problem but do it quicker _and_ can make a profit. The cost
of a basic reflective sunshield would be astronomical and
would provide no financial return whatsoever. Not only that
but we don't have the technology to bring the millions of
tonnes needed into Earth orbit let alone make a shield out
of it.

True and not true. There is one thing you have forgotten. In 1815
Tambora erupted and we had the "year without a summer". A space based
system would be capable of INCEASING the amount of sunlight falling on
Earth. In the next volcanic eruption we might well try to persuade
people to burn MORE carbon!


- Ian Parker

"Ian Parker" wrote in message
oups.com...
On 23 Feb, 15:21, "George Dishman" wrote:

True - however the time constants for mathane etc. are very long.
Longer in fact than the timescales for te chnological development.

No, that's the point. CO2 timescales are long - the
heating effect of what is already in the atmosphere
will continue into the next century, but methane is
short lived. nethane burns in air, CO2 doesn't, and
it will react with the oxygen and be destroyed
comparatively rapidly. Go back and read the original
article again.

The time constant is caused by the fact that there is methane in
permafrost. As this thaws methane gets released. There has been
speculation about the cause of ice ages. The simnple fact may well be
the unstable equilibrium caused by methane.

Yes, that's the main unknown, the point at which the
positive feedback becomes runaway.

The fact that methane
oxidises means that it has a finite life, but that lifetime (in lower
than ignition quantities) may not be that short.

Biology produces large amounts of methane (think of
cattle) and the lifetime is already well known.

Look - I may be prepared to concede some points on timescales, but AI
MUST be the key. Look if you can send a robot with all the
manipulative properties of an astronaut, ...

Again, that's the point, we don't have robots with
anyhting like those manipulative abilities.

I will comment on this with general AI comments.

.. you MUST be able to have a
robotic colony. If AI does NOT produce a VN swarm then astrnauts MUST
be a wate of time too. I mentioned Sonic in particular because his
software is going to be used by the Pentagon in mititary robots. I
find it interesting too that military technology is no longer the
leadimng technology. Games and animation seem to be.

I picked up on your mention of Sonic as mostly tongue-
in-cheek but it was intended to emphasise the fact that
no amount of intelligent software will get round that
manipulation problem.

Being able to do somthing, no matter what it is, is a matter of being
able to predict the consequences of actions. This is why games are
relevant.

No, being able to do something depends on connecting
with the physical world.

Let us do a little thinking round this problem Suppose we built a
robot with similar joints to a human and with electric motors or
fluidic circuits to activate them. Personally I like the idea of a
fluidic robot woth one central pump and pistons for limbs.

Sure, that would be my prefernce too. Call it a JCB.

Building
this (without AI) is a fairly trivial extercise. We could provide
pressure sensors for the finger tips to provide a general "feel".
There is nothing here, except of course for the AI element which could
not be built by a reasonable tean of engineering undergraduates.

Such designs have been done but their dexterity is very
limited. There is a huge market for prosthetic limbs that
can be addressed by such devices, so if they could be
"built by a reasonable tea[m] of engineering undergraduates"
why aren't there a lot of wealthy engineering undergraduates
around?

Of course without a brain our assembly could not do anything useful.
If it had radio transmittor/receivers it could do a range of things
with telepresence but it would not be capable of acting autonomously.
Put in AI and it then becomes an autonomous machine that you could
send up to repair Hubble, and if in space would do everything an
astronaut could do - and more as we know that robots in medicine are
capable of cutting with a precision that no surgeon can manage.

Sure it could, and if we had anything like that capability
it would be up there doing it, but nobody has built a
general purpose robot hand that can put a nut on a screw
with sufficient feeling to avoid cross-threading it.

The point is that there are economically viable processes
that we can use here on Earth which will not only solve the
problem but do it quicker _and_ can make a profit. The cost
of a basic reflective sunshield would be astronomical and
would provide no financial return whatsoever. Not only that
but we don't have the technology to bring the millions of
tonnes needed into Earth orbit let alone make a shield out
of it.

True and not true.

True. The alternatives I have suggested may not yet be at
break-even point but all provide a commercial return which
is comparable to the outlay. Your scheme is vastly more
expensive, thousands of billions of dollars at least, and
provides no return whatsoever. It is not a credible
alternative to existing schemes in any way whatsoever.

There is one thing you have forgotten. In 1815
Tambora erupted and we had the "year without a summer". A space based
system would be capable of INCEASING the amount of sunlight falling on
Earth. In the next volcanic eruption we might well try to persuade
people to burn MORE carbon!

That comment does not address the economic case I made above,
nor is it even correct.

We have has several volcanic eruptions in the past few decades
and while their effect is visible in the records, it is small
compare to the greenhouse heating and lasts only a few years.
The CO2 already in the atmosphere _now_ will still be causing
elevated temperatures next century so while a major eruption
might give some partial temporary respite, its effects will
disappear in less than a decade. Nobody will be trying to burn
more carbon until the fossil fuels have run out and we are
looking for a way to combat the next ice age, and that could
be thousands of years away.

George

On 1 Mar, 23:29, "George Dishman" wrote:
"Ian Parker" wrote in message

The time constant is caused by the fact that there is methane in
permafrost. As this thaws methane gets released. There has been
speculation about the cause of ice ages. The simnple fact may well be
the unstable equilibrium caused by methane.

Yes, that's the main unknown, the point at which the
positive feedback becomes runaway.

We could be at that point now
Let us do a little thinking round this problem Suppose we built a
robot with similar joints to a human and with electric motors or
fluidic circuits to activate them. Personally I like the idea of a
fluidic robot woth one central pump and pistons for limbs.

Sure, that would be my prefernce too. Call it a JCB.

Building
this (without AI) is a fairly trivial extercise. We could provide
pressure sensors for the finger tips to provide a general "feel".
There is nothing here, except of course for the AI element which could
not be built by a reasonable tean of engineering undergraduates.

Such designs have been done but their dexterity is very
limited. There is a huge market for prosthetic limbs that
can be addressed by such devices, so if they could be
"built by a reasonable tea[m] of engineering undergraduates"
why aren't there a lot of wealthy engineering undergraduates
around?

Good point. Dexterity is a function of 2 things. It is a function of
the AI system and it is also a function of the motors used. Prothetic
limbs up to now have used stepping motors which stop after every
click. It is this insistance on stopping after every click which is
limiting dexterity. An agile Prothsetic Limb would not have this
limitation. Indeed we could define "agile" a meaning unable to stop
after every click. It is interesting to note that when normal limbs
move the die is cast as soon as they start moving. Indeed sports
coaches have just discovered this and they tell a sportsperson to look
elewhere as soon as the "commit" is over.

The other thing an APL needs is a measure of AI. If the nerves can be
grafted on this problem is circumvented. An APL could be controlled in
a number of ways. It could be controlled by balance, an artificial
horizon in a pair of spectacles could ensure that you did not fall
over and balance was maintained. Full AI could take you up and down
stairs using a camera.

Stepping motors are used to some extent because of the lack of
confidence in AI. With AI one would immediately abandon the need to
stop.

There is one point on social policy and cost. Would it be cheaper to
develop an APL or to pull down buildings and use ramps instead of
steps. Using ramps strikes me as being very expensive. A footbridge
with a ramp costs many times one with steps. In fact to simply climb
steps you do not need full AI. You just need a set of LEDs (possible
NIR) on each step as APL aligners.

Another point. You say later on that the cost would be prohibitive. I
think the cost of ramps is prohibitive! There is an important point
here. If this is done in stages there will be money all the way down
the line. NOT the case with ISS.
Of course without a brain our assembly could not do anything useful.
If it had radio transmittor/receivers it could do a range of things
with telepresence but it would not be capable of acting autonomously.
Put in AI and it then becomes an autonomous machine that you could
send up to repair Hubble, and if in space would do everything an
astronaut could do - and more as we know that robots in medicine are
capable of cutting with a precision that no surgeon can manage.

Sure it could, and if we had anything like that capability
it would be up there doing it, but nobody has built a
general purpose robot hand that can put a nut on a screw
with sufficient feeling to avoid cross-threading it.

This is being developed by Andrew Ng at Stanford. When you say "it
would be put up there". The establishment of NASA is wedded to manned
space flight. Hubble could have been designed differently. Good idea -
cross threading is a problem we might give a student!

You are raising a number of specific problems. I don't believe these
are showstoppers. They do mean that you need a completely dedicated
team. Andrew Ng is also doing research in linguistic AI. BTW - Google
is going into AI but they don't seem to know what they are about.
Stations are still elastic "La estacion de resorte". Whle I wonder
whether you would want to go out with a "fosforo". It does not seem to
know the different types of match.

True and not true.

True. The alternatives I have suggested may not yet be at
break-even point but all provide a commercial return which
is comparable to the outlay. Your scheme is vastly more
expensive, thousands of billions of dollars at least, and
provides no return whatsoever. It is not a credible
alternative to existing schemes in any way whatsoever.

I think the intermediate points will.
There is one thing you have forgotten. In 1815
Tambora erupted and we had the "year without a summer". A space based
system would be capable of INCEASING the amount of sunlight falling on
Earth. In the next volcanic eruption we might well try to persuade
people to burn MORE carbon!

That comment does not address the economic case I made above,
nor is it even correct.

We have has several volcanic eruptions in the past few decades
and while their effect is visible in the records, it is small
compare to the greenhouse heating and lasts only a few years.
The CO2 already in the atmosphere _now_ will still be causing
elevated temperatures next century so while a major eruption
might give some partial temporary respite, its effects will
disappear in less than a decade. Nobody will be trying to burn
more carbon until the fossil fuels have run out and we are
looking for a way to combat the next ice age, and that could
be thousands of years away.

Tamborra was the biggest. True you don't want more CO2 but you might
want more sunlight.

People are scaring us with a VEI 8 such as Yellowstone.


- Ian Parker