Let's suppose that carbon-based CHON life is indeed the only likely path to naturally evolved life. Let's further suppose that an oxygen atmosphere and aerobic life is the only likely path to complex technological naturally evolved life.

Even so, aliens might not be limited to "naturally evolved" life, and they might not even be interested in "naturally evolved" life forms. They may be artificial life forms.

What would artificial life be based on? Well, if we humans were making them, then they'd likely be based on silicon, iron, and aluminum. Hydrogen would not be a requirement.

Now look at this "lava planet", COROT-7b:

http://www.space.com/scienceastronom...rock-rain.html

This planet's likely similar to Mercury on steroids. No hydrogen, but lots of really hot rock--so hot that it has a "rock atmosphere". Obviously, CHON life as we know it has no use for this planet, but what about silicon/steel life? It has oodles of raw resources in easily aeroscooped gaseous form, there's oodles of solar power, there's a deep gravity well for Oberth effect maneuvers, there's an atmosphere for aeromaneuvers...it's pretty awesome.

That got me thinking--COROT-7b is pretty extreme and perhaps rare, but hot jupiters are common and easily detected. What if we assume most hot jupiters are surrounded by extensive moon systems? These moons would be similar to COROT-7b; lots of concentrated metals. And while they'd likely have no atmosphere to work with, the hot jupiter itself would have an atmosphere and gravity well to make for convenient Oberth effect and aeromaneuvers.

It seems to me that hot jupiter systems could be prime locations for artificial life to thrive. There are awesome levels of the right sort of resources and power. The location is convenient for getting around. They're easy to detect from far away. Who could ask for more?

Now let's look at our Solar System--from this perspective, it looks horrible. There's no hot jupiter system. The deep inner system is a wasteland; there's only little Mercury and it has no atmosphere to speak of. The outer solar system has little solar power, and the desired resources are diluted with a lot of ice.

Maybe the aliens are perfectly aware of the existence of our Solar System and its planetary system, but they've dismissed it as just a minor footnote in their star catalogs. Nothing useful to speak of, so no reason to visit or investigate further.

Hot jupiters would have smaller Hill spheres, because the size of the Hill sphere is dependent on the distance between the planet and the star as well as on the mass. A smaller Hill sphere means fewer moons.

Most of the mass making up the moons of the gas giants in our system is probably ice; the only really rocky outer moon is Io. Ices would not persist on the moons of hot jupiters so that would also cut down on the available mass in orbit around gas giants in general. So it looks quite likely that hot jupiter moons are reasonably rare.

But the ones that do exist would, as you say, be valuable real estate.

__________________
New Orion's Arm Site . The Starlark . Against a Diamond Sky (OA Novella Collection) . OA Flickr set

I'm pretty sure as well that self-replicating and (hopefully) intelligent artificial life will use quite a lot of CHON in its construction. Carbon is just such a versatle element that I can hardly imagine it having no use in making future high-tech devices. Trying to make a self-replicating system using without carbon, oxygen and hydrogen would be very difficult. Nitrogen might be useful too. Iron, aluminium, silicon and others would be useful, but not by themselves.

__________________
New Orion's Arm Site . The Starlark . Against a Diamond Sky (OA Novella Collection) . OA Flickr set

We don't know that for sure, but in any case fewer moons than, say, our own Jupiter system still means a lot of moons. They might even be Earth sized.
And as I understand it, this is because Io is close enough to Jupiter that the initially hot Jupiter (hot due to gravitational compression) radiated enough light to blow away its volatiles (just as the Sun blew away Mercury's volatiles).
Perhaps it cuts down on the overall mass, but what's left are the silicates and metals that artificial life forms may desire anyway.
Perhaps, but I wouldn't be surprised if they don't use hydrogen much. The basic problem with hydrogen is that it's mostly available where solar power isn't, or where it's deep in a gas giant's gravity well.

Artificial robotic life could do just fine without utilizing hydrogen. I don't see a compelling reason to discard oxygen, since silicates and oxides should be pretty plentiful (although nickel-iron asteroids/moons may be useful in their own right).

But I'd try to keep things simple rather than try to exploit everything out there. In principle, self-replicating robots could use nothing but crudely refined iron wires and arbitrary oxide powder for electrical insulation. This assumes crude electromagnetic relays for logic circuits and dynamos for electrical power generation.

Here's a very basic step on the way to self-replicating devices; the RepRap
http://en.wikipedia.org/wiki/RepRap
Currently they are concentrating on thermoplastic polymers, as they seem to be easiest to work with. Hydrogen is required for that kind of material.

Perhaps a mature self-replicating technology would use lots of different recipes, one suitable for use on the moons of hot jupiters (if such exist), one suitable for use in Earth-like temperatures, one suitable for use on Venus-like worlds, and so on.

If it is possible to make self-replicators that dont require hydrogen, then a small inner world of any description would be a good target and source of materials. When we finally get to examine Mercury's surface perhaps we'll find a population of alien von Neumann machines, happily mining away.

__________________
New Orion's Arm Site . The Starlark . Against a Diamond Sky (OA Novella Collection) . OA Flickr set

Prototyping in plastic makes sense here on Earth, where we have a robust hydrocarbon economy, but even so there are problems. RepRap needs all sorts of metal components, including wiring for motors, sensors, and control electronics.

The bottom line is that you can't build self replicating machines with just plastic, but you can build them with just metal and ore (metal oxides).

Now, using lots of different recipes may be more...hmm..."comprehensive" than using just one or two recipes, but it's not necessary and whether it's "desirable" by the aliens depends on their goals and mentality. Do they want to exploit every last cubic centimeter of real estate out there? Just because they can, perhaps? Or do they just want to inhabit places which they find pleasant or aesthetically pleasing? The room for speculation is as big as the universe, here.

My understanding is that Europa is almost entirely rock (water/ice shell is no more than 100 km thick), while Ganymede and Callisto have much thicker ice shells, but still about as much rock as either Io or Europa. Evaporate the ice, and they all would be about the same size.

__________________
Fiction has to be plausible. Reality is under no such constraint.

Say you were able to accurately model the energy budgets of particular enviroments, such as an extra-solar planet or icy moon, could one tell, perhaps if the radiative budget was not being balanced, that the environment was harbouring life, or some similar process. In a simple case, where the losses of heat to space balanced heat from the core and the solar flux, would the presence of photosynthetic life absorbing sunlight and reproducing, produce a thermal signal?

__________________
plenty of woo, at the hotel hoagaland...

IsaacKuo, I don't really understand your thought that a Machine civilization would not want or use Hydrogen. What about nuclear Fusion? Or using it as spacecraft propellant?

Surely any advanced civilization, Organic or Mechanical, would want to use the most abundant resource in the universe. And as for Carbon, what about carbon nanotube structures, etc? Surely they would be useful to a race of machines in any number of possible scenarios, building space-elevators, surviving extreme pressures and so on.

__________________
The Sky is no longer the Limit

I don't think so. Whatever energetic activity is being powered by the absorbed sunlight, it will mostly end up converted into waste heat. That will look more or less the same whether the sunlight is absorbed by solar panels, photosynthetic plants, or lifeless rocks.
My idea is that hydrogen actually isn't abundant, where you'd want it to be. Generally, you want to be where there is plentiful solar power, and this tends to be where hydrogen rich volatiles have already been blown away by sunlight. Deep gravity wells can save hydrogen from this fate, but this puts those hydrogen resources deep within gravity wells where it's expensive to get.

If you've mastered fusion power, then that might change things--you may not care about solar power if you can get all the power you need from fusion reactors. In this case, cold jupiter systems like those in our own Solar System may be ideal. However, I am skeptical of the practicality of fusion power technology.
Carbon may be a generally useful supermaterial, especially if it's generally available in CO2 heavy atmospheres (as in Venus and Mars). It may also be useful in making steel, and it's possible carbon may be a generally superior choice for building electronics, as opposed to silicon. We're living on a conveniently silicon-rich world, so that may have biased our technological development.

But even if carbon isn't so generally available, there's no shortage of alternatives.

I think it's pretty much a given that any advanced spacefaring civilization will have mastered fusion. Why are you so sceptical about it? We're on the verge of mastering it on Earth within the next few decades. Whether it's in Tokamaks or by Laser, Fusion is a reality already. And even if you couldn’t for some reason master controlled fusion, there's always uncontrolled, i.e. Hydrogen bombs. For a civilization of machines which don't mind radiation, you might be able to get power by letting off H-bombs and absorbing the power.

And as for the Hydrogen "not being where you want it", that assumes that the machines can't live in a Gas Giant's atmosphere, which I think they probably could. It might be an ideal environment for them.

You seem to be limiting the possibility for a machine civilization to only the types of machines and technologies we have today, it seems rather Steampunk.

__________________
The Sky is no longer the Limit

Because it's not at all clear that practical fusion power technology is even possible.
Yeah, right. We were a few decades away from mastering it a few decades ago. Except we weren't.

We have indeed mastered fusion in the sense that we understand it very well and can reproduce fusion reactions in the lab at will. But we do so by putting more energy in than we can get out. So, we have not mastered fusion power technology. And perhaps we never will.
H-bombs require large amounts of weapons grade fissionables, which makes it very expensive. Then there's the technical challenge of storing/converting the energy generated into useful forms. One method, for example, might be to set off the bomb in a large cave to heat water and then run steam turbines...but if you do that, then you might as well do geothermal energy, which is easier and cheaper.
Living in a gas giant atmosphere would be quite a challenge. The basic problem is that there's a lot of gravity that wants to pull you down to the core. You could try to use a balloon, but what do you use for a lifting gas? There's nothing lighter than hydrogen, and that's what the atmosphere is made of. Or you could use powered flight, which means you spend a lot of power just staying aloft. And you have to deal with strong winds and convection currents that are also trying to rip you apart and pull you down to your doom.
I gave the example of how we'd build robots as a concrete example. Certainly there are other possibilities, but my main point is that CHON isn't even the most likely for robots designed by us.

It's clear that fusion technology is more possible than the technology you propose in your opening post. I'm astounded that you feel it would be beyond the capabilities of a civilization that would be capable of other feats you would think them capable of.

Really? How do you design a fusion power reactor, then? No one knows how.

What capabilities have I described in the opening post which you find surprising? Solar panels? Silicon based electronic circuitry? The Oberth effect and aeromaneuvers? Obviously not--these are things which we have today.

The only novel technologies I notice are aeroscoop technology and self-replicating machines. The closest thing we've flown to a true aeroscoop would be the Stardust mission. And while we have implemented programmable prototyping machines, none of them so far have the ability to manufacture itself. But there's nothing to suggest it's impossible on the grounds of physics or engineering limitations. Not like fusion power, which runs up against a lot of hard inescapable physics constraints.

You do know that ITER is expected to achieve Breakeven right? Operations start in 2018 and will last for 20 years (assuming it's all on schedule). But even if ITER and DEMO (after that) don't achieve useable power, there's the Laser confinement option and potentially others in the future. Whether we get Fusion in the next few decades or whether it takes longer, we will have it some time.

Then why is there tens of Billions of dollars put into Fusion research and practically nothing put into self replicating machines? Because Fusion will actually be practical and useful within this century.

__________________
The Sky is no longer the Limit

"....artificial life forms."

All this is says is that it MIGHT happen. Maybe.
No, its because fusion RESEARCH is practical and useful, and it's because the fusion research industry has awesomely good P.R. for getting funding. Fusion research is scientifically fruitful even if it never produces a fusion power reactor. The potential for developing fusion power is certainly a good thing to get politicians on board, but it's not the whole story.

Magnetic confinement fusion research has led to a lot of excellent plasma science and may produce fruit in the form of advanced propulsion technologies like VASIMR. Laser inertial confinement fusion research may lead to advances in high power laser technology, perhaps leading to the development of laser powered SSTOs, weapons grade lasers, and so on.

If fusion power were a sure thing, then we wouldn't need government funding to get involved. Private investors would be clamoring to buy in on the ground level. But they aren't, and it's for simple business reasons.

Self replicating machines are a technology which we simply don't need with any urgency. We can already build things with non-self replicating machine technology, so what's the rush? Even so, there's research in prototyping technology which will provide self-replicating machine technology as an incidental side effect. It won't cause anything like an economic revolution, because it's still cheaper to build things using traditional mass production techniques than with prototyping machines.

Self replicating machines will be a largely uninteresting niche side effect of a much larger industry in prototyping machines.

Fusion research will continue to be a broad area of research which produces useful side effects even if its main political "hook" never bears fruit.

Pretty different!

At least we know self-replicating things are possible. They occur naturally, so there's no physics limitation which rules it out.

In contrast, there's nothing even remotely close to a practical fusion reactor out there. Stars exist in nature, of course, but they're mind numbingly slow. Even in the hot pressure cookers of stellar cores, it takes billions of years for a proton to react! Consider if you designed a coal plant like that--you throw in a coal pellet and you have to wait billions of years before it burns.

"Self-replicating things" are a far cry from "....artificial life forms."

And fusion bombs are mind numbingly fast. Somewhere between the uncontroled fusion of hydrogen bombs and the unsustained fusion of experimental reactors is a middle ground that should be achievable. I'm surprised you don't think so.

The problem Isaac is trying to address is one that is a real one. If you send a space probe filled with self-replicating machines to another star, where do you send it so that it has the best chance of replicating? If you send it to the inner system there is a lot of energy for replication, but very little hydrogen; if you send it to the outer system there is a lot of hydrogen and not much energy.

In the inner system a self-replicator would probably start by growing photovoltaic cells or some other sort of solar panel and would rely on metals and semiconductors. In the outer system the self-reppers would have to try to build a fusion generator from scratch (unless they brought one with them) or try to use tidal or magnetic energy, but they could use plastics and organics (which may perform certain self-replicating tasks well).

Neither prospect is ideal.

A third option - self-replicating on an Earth-like terrestrial, with plenty of C,H,O, and N, is appealing, until you realise that your self-reppers are trapped on that planet until they can build a launcher capable of reaching escape velocity. Earth-like worlds might be rare, and they are always difficult to launch from, but they do offer plenty of exploitable resources.

__________________
New Orion's Arm Site . The Starlark . Against a Diamond Sky (OA Novella Collection) . OA Flickr set

What's the difference? Consider the definition of natural "life" before answering.
A fusion bomb can't just be put into slow motion, and current fusion research reactors can't just be made magically more efficient. They work in fundamentally different ways; there's no continuum between them and as such there's no a priori reason to expect a "middle ground" between them to be achievable.

Fusion bombs aren't really "fusion" bombs anyway--they're more fundamentally fission/fusion bombs. No fission, no fusion. No fusion, no U238 fission. Large amounts of U238 are a waste produce of U235/plutonium refinement, so the extra U238 fission is a huge bonus.

Fissionables, of course, can be used for nuclear reactors. It would be really nice if fusion could be used to augment atomic reactors in an analogous way as they augment atom bombs, but there's no particular reason to believe it's possible. The mechanisms involved are too different and practically unrelated.

Is an artificial replicator life? For that matter, is a natural replicator life? If you were to gather the prominent biologists of the world in a room and have them hash out that question, they may well come to the conclusion that viruses are not living, just as a group of astronomers recently concluded that Pluto is not a planet. In both cases there is much debate and there will never be total agreement on the definitions.

Even if we consider viruses life, and so might consider an artificial replicator life, there remains the question of how easy it would be to make one. Making nano machines is tough enough. Making nano replicators that could be considered life enters the unknown.

I guess its a matter of opinion, but I see the technical problems involved with making a fusion reactor as more straightforward and so more easily addressed.

Whether in the center of a star, a bomb, or a reactor, the fundamentals are the same: Heat/pressure and fuel mix. There is a continuum for each of these variables. The mechanics for achieving the proper combination may be very difficult, but I do not see why you would say there is some physical law forbidding it. We are simply talking about heat/pressure and fuel mix, not time travel.

Yes.
A virus is not traditionally considered life because it is not a natural replicator. It needs living cells to do the work of manufacturing more viruses. It's like a blueprint or computer program.

You wouldn't call a computer program a computer, because it just isn't. It might be part of a functioning computer, but it would at best be a component of a computer.
Really? Then how do you make a working fusion power reactor?
No, it isn't.

As I noted, so-called fusion bombs are actually fission-fusion bombs. The way they operate is fundamentally intertwined with fission, and as such they depend on fission principles. The big factor in fission is neutron manipulation. It's largely about generating and manipulating neutrons to be in the right place with the right energy levels. The fusion portion of a fusion bomb can be largely thought of as a neutron generator; it's the massive flux of generated neutrons which allows fissioning U238 for the final stage of energy generation.

The design of bombs and reactors is more complex and subtle than you're thinking.
We can get the right heat/pressure mix to generate fusion. It doesn't even take a very complex or expensive device--the fusor is a straightforward way to do it with high school science lab level hardware.

The challenge is to come up with a reactor design which can generate power, and a fusor can't do it. The laws of physics prevent it from doing so, due to various issues with maximum charge density and losses to impacts with the grid elements. It MIGHT be possible to design a successful fusion reactor using a similar principle to the fusor...but maybe not.

I see life as being much more complex and subtle than a fusion reactor. Sure, a fusion reactor is a long way off and the difficulties are great. But the difficulties of creating artificial life are unknown.

It is very interesting to try and find the dividing line between life and just a very complex machine. I suspect it has something to do with thermodynamics and information dynamics. Life can manipulate energy and information in ways that non-living systems can not. I'm not smart enough to understand the details, but I'm smart enough to know they are very complex and, I suspect, still largely unknown.

How would you design an artificial life form?

I see it the opposite way. We do not know whether a fusion power reactor is even possible. We do know that natural life forms actually exist, so they are indeed possible.

As for the complexity of implementing a self-replicating machine...there's no reason why it would be anything like the complexity of naturally occuring life forms. Compare a chess playing computer against the human mind, the only naturally occuring computer capable of playing chess. Any way you look at it, the natural solution is more complex than the artificial solution.

Similarly, there's no reason why an artificial self-replicating machine should be as complex as a bacterium. Even the simplest of known bacteria involve myriad complex molecules, which make a typical non-biotech chemical factory look simplistic.
I would keep it simple. That's why I have advocated in this thread to keep the raw materials down to a minimum. Instead of cleverly trying to exploit all materials, I say design the self-replicating machines to only use a few raw materials, like iron and crushed rock (for the insulators).

I think an artificial self replicating machine should be relatively simple, not relatively complex.

As for what level of "intelligence" is appropriate for these machines--well, if the machines have "human level" smarts, then I suppose they'd be clever enough to expand beyond their original programming and they could exploit CHON and such...but I don't view this as necessary nor even necessarily desirable. "Intelligence" on the level of a chess playing program may be good enough, and may even be all we really wish to imbue onto our machines anyway. Give them enough smarts to seek out particular planetary systems with particular resources that they're programmed to exploit. Sounds good to me.

Now I remember in the Sept 09 PopMech, that there was this "walking router robot" that might be even more useful. Alas I could not find it here:
http://www.popularmechanics.com/mythbusters/
http://www.popularmechanics.com/auto...html?series=56

Ah! Here it is http://www.hexapodrobot.com/index.html http://hacknmod.com/hack/incredible-...ter-cnc-robot/


We may be seeing these robots soon
http://www.popsci.com/technology/art...olution-bigdog

http://www.popularmechanics.com/blog...s/1764167.html
http://www.popularmechanics.com/scie...e/1764027.html

I think you are drastically underestimating the complexity of the problem. Not only does a self replicating machine need to reproduce its mechanical system, it needs to reproduce its power system, sensory system, and its information system. If its information system is a computer, it needs to reproduce that computer.

I think the requirements will become very complex very fast. What do you see as the absolute simplest system possible?

It depends, somewhat, on what the criteria for "simple" are. For me, the big criteria is simplifying the "supply stream". The basic thing that these self-replicating machines need to do is "feed" on raw supply material in order to do the rest of these things. It does no good to "simplify" the process of constructing the machines by assuming a supply of independently manufactured servos and microcontrollers, etc.

(For a business model here on Earth, such a supply of servos and microcontrollers is a feature, not a flaw. How are you supposed to make a profit if you don't have supporting "stuff" that the customer needs to keep on buying?)

So that's why I propose using iron and oxide ore (which is where you get the iron from). Iron is a very flexible material. Not only can it be used for structural mechanics, it can be used for electronic wiring and components. Iron can also be magnetized, which is important for creating a dynamo (which is where you get electrical power from--hook up some sort of turbine to an electric generator).

Mechanics? Obviously, iron can be fashioned into sturdy hulls, struts, gears, wheels, and such. Importantly, iron can be fashioned into wire coils for electric motors. Note that you need an insulator material also; I suggest oxide ores because you need them anyway to get the iron.

Power? Iron can be fashioned into coils and magnets, which is the basis for electric generators. If there's an atmosphere to work with, you can make wind turbines. In deep space, you can make solar thermal turbines--you need some sort of working fluid; oxygen liberated from oxides during iron mining/refining is an acceptable working fluid.

Sensors? At a crude level, iron can be used for simple switch based touch sensors. It can also be used to make coils and magnets for sonic transducers--giving you ultrasonic sensors. For deep space sensing in a vacuum, you can use old fashioned vacuum tube style TV camera technology for visual sensing.

Information system? Vacuum tube style electronic logic is possible, as is more simplistic electromagnetic relay digital logic (the switches are relays rather than vacuum tubes or transistors). Relays may be less efficient than vacuum tubes, but for purposes of "simplicity" they may be better--they use the same basic components as the motors/generators.

Is all of this complex? Sure. But the key for a proof-of-concept, I think, is to make sure that it can be built from a minimum set of basic components.

Now, I don't think you'd strictly use this criteria for designing an actual mission. While electromagnetic relay logic is the "simplest" solution, it would not be the most "effective" solution. The additional hardware and infrastructure required for silicon logic chip manufacture may be worth the extra effort and complexity.

Vacuum tube technology! Sounds like 1950's SiFi.

"...additional hardware and infrastructure required for silicon logic chip manufacture..."? And all that needs to be reproduced from scratch as well. Impossible any time in the foreseeable future.

Do you think it is possible to construct a vacuum tube (or silicon chip) using only iron and silica? And using equipment made of only iron and silica? Any additional materials will need to be found and extracted from the environment, requiring specialized sensors that also need to be replicated and probably need additional elements for their construction that require additional sensors to find that must also be replicated. How do you avoid getting onto this spiral of complexity?

It is indeed a spiral of complexity, but it could be worth it because of two factors:

1) We have a lot of experience with silicon microchip based computers/robotics/sensors/etc.

and

2) The performance benefits of microchip transistor based electronics over "simpler" technologies are so great.

Factor 1 means that we can leverage our existing microchip designs and software for the design of these self-replicating machines. We wouldn't have to design everything from scratch.

Factor 2 means that we can't just "emulate" familiar microchips and software using alternative technology. For instance, if we just use electromagnetic relays, then even a 6502 class processor would be as big as a refrigerator and would be about a thousand times slower.

And it's certainly NOT possible to construct a silicon chip with just iron and silica. Semiconductor transistor technology requires doping with other elements. The required amounts may be relatively small, but you still need to get them from somewhere.

Vacuum tube technology may be more practical, because all you really need is iron, vacuum, and some insulative structural base. But even so, vacuum tubes are inherently less efficient than transistors and the performance of a "vacuum tube chip" will be inferior to a semiconductor chip.