Grey goo scenarios usually violate the second law of thermodynamics. They might be fun, but they're not hard sci-fi. Even the ones that do not tend to run a thousand to a million times faster than they could.

Nuclear batteries seem to be common. They will of course be made extremely resilient, but occasionally somebody manages to railgun one of them. Similar for all the other radioactive materials used, from nuclear fuel to the exotic isotopes antimatter production probably produces: they can get spread out in a habitat, giving all the biomorphs lethal radiation. Sure, it is less deadly than it was (bananas furioso, resleeving), but nobody likes puking their morph's guts out while waiting for their turn in the egobridge. Collisions in space are bad news. Ships (and debris) travel faster than most railgun bullets. If a ship hits your habitat it will wreck it in a hypervelocity explosion. This is why most habitats take traffic control dead seriously, use tugs to guide ships to dock, and have heavy batteries to deal with incoming objects. Ship engines have already been mentioned. It is worth noting that a fusion or antimatter drive is essentially a particle beam: the destroyer can (if my calculations were right) melt through hundreds of meters of matter. Ouch. Metallic hydrogen is very explosive: it wants to be a light and fluffy gas. When expanded (at NTP), it takes up about 7000 times as much space: boom! This means that a ship tank, refuelling station or vehicle with MH can explode nicely. And if they do it in an oxygenated environment the hydrogen also reacts with the oxygen, creating an even bigger boom. Habitats and ships that rotate are subjected to complex torques from the Sun or nearby planets, as well as internal changes as people, cargo and water moves around. In order to manage them they have flywheels and gyros that compensate. If they fail badly they can explode - or create a wrench that breaks the habitat. While naive grey goo is implausible (most stuff is at the lowest energy level, so in order to turn it into grey goo you need energy) self-replicating nasties are plausible. Solar-powered black goo might act as a fast lichen, growing over everything. Nanomachines powered by the energy content of biomass might work, burning it into more machines. Little gremlin-bots copying themselves can overrun a habitat like vermin, draining resources and causing equipment failures. Nanoplagues and bioplagues are all to real. Not to mention a new computer virus that there is no antiviral software against: suddenly the whole local mesh is gone.

Static DET5 wrote:

Would you be willing to walk me through this? If we accept the idea that practical nano-fabrication is a reality, then how is gray goo not a real threat? I know that the process will take energy, but society is teeming with energy. Habitats spend large amounts of resources generating and diffusing energy. Granted, an ultra-simple grey goo isn't going to preserve any native energy sources, but there will still be solar flux to help poke things along. Hell, the grey goo disrupting the local power sources constitutes a grey goo industrial accident, almost by definition.

If the grey goo taps into the power mains it can increase its energetic potential, sure. But that is easily detected (I'm sure there's hundreds of millions of sensors on the power grid just measuring power draw) and remedied (send in the counter-swarm). As for solar, if you assume the tech can get up to 40-50% efficiency (4.5x photosynthesis) and zero of the energy is dedicated to support of the solar collection, you're beating earth plants by about 10x. 10x of a plant is more like a super obnoxious weed. So in terms of industrial disaster, sure, but to compare to adjectives like "horror and danger" "terrible and habitat-destroying" its more like "frustration and annoyance" and "budget-breakingly expensive to remediate and so obnoxious some people move somewhere else eventually." I suppose it counts but its not very dramatic. EDIT: Re: solar - solar flux is about 2.25 times less on Mars, and its possible to turn off the lights on a habitat (even the outside - tug the hab to a perpetual night zone or put up a shade). EDIT2: If world spanning solar powered grey goo is feasible, then honestly Mars should already be covered in it. That would be a fantastic way to speed up the terraforming of Mars. It's not though.

On one hand, we have nanotech that is capable of fabricating whole goods in a practical amount of time. On another hand we have folks saying that grey goo is a physical impracticality. I still haven't heard why grey goo wouldn't be considered an industrial accident, potentially of epic proportions. This is one of those "penny problems". On day one you put a single penny in the first square of a checker board. On day two, in the next square, you put two. Day three, next, four pennies. Etc. Actually, I'll change the analogy slightly, to an e. coli bacterium. At replication interval one you go from one bacterium to two. The interval is 20 minutes. Every 20 minutes the population doubles, right? We get a (2^n)=P equation, where n= the number of generations, and P equals the final population. Someone check my math, but by the end of the day we're looking at more than 5*10^216... That's insane tons of bacteria. Granted, we don't see this happen because of external forces on the bacteria, but the bacteria are also incredibly limited on what they can metabolize and generate new bacteria from. What's the generational time on a self-replicating nano-bot? Instead, we're talking about an industrial system that is apparently designed to do nothing but break things down to the components necessary to build more of itself. It is able to harness energy at a higher efficiency than the biological system (you've allowed for that in your musings). Hell, let's look at it this way: The Earth is infested with bizarre green stuff (plants). It developed almost spontaneously, and over several millions of years it diversified across the planet until it infested every square inch of the surface. Remember, we're not trying to destroy everything. We're talking industrial accident.

It makes sense to self-replicate if that's all (or almost all) that you are after. Put a swarm out there, and it starts making more. It may be a system of nanobots responsible for making new nanobots, but short of hand-waving and saying "It just isn't possible", I don't see why it isn't possible (P.328 of Eclipse Phase does do this hand-waving). I'm not real good at handling the hand-waving, particularly given the other state of nano-technology. Given that Proteans exist (P. 329, EP), it seems a short step to a community of Proteans being designed to build more (Again, why? I admit, I can't answer that one). At first blush, computer viruses don't make sense. They're inconvenient (just wipe the affected computer, right), they're limited (until recently they affected one "environment" they were programmed for), and they typically don't result in real gain by the creator (besides notoriety, offset by criminality). Why would anyone create these? Initially, it was because they could. The EP current state of nano-technology allows for the creation of 10m^3 orbital transfer vehicles through the use of nano-fabrication (P. 327, EP). This is a massive volume. For comparison, I offer a spreadsheet: https://docs.google.com/folder/d/0B7Qo7h9CAigaYkFBU1lsc0VtMTQ/edit?usp=s... The sheet is a very simple, quick and dirty effort to calculate molecular construction by nanobots given immediate supply and energy. The variables look at the number of molecules to move, the nanobots moving them, the distance to move them, and the speed at which they move. The actual sheet is on page two. Page one is just a metric conversion scale that I find useful. The sheet is configured for the manufacture of an extremely simple product, the bullet (the projectile) for a US Army .45ACP lead bullet. Here's a quick snippet from my notes: It would take a single nanobot, moving at the speed of light, over 679 years to build a .45ACP lead bullet (not round), if the construction head were 1 millimeter away from the build location. This assumes perfect build conditions and no loss/maintenance of nanobots. One billion nanobots can achieve this in two and a half minutes. One trillion bots(1*10^12 or a tera prefix) can drop this time to 15 hundredths of a second. Taking those trillion bots and slowing them down to 10 meters per second (A fast human run), and we're back to about 7.5 weeks. Increasing the nanobots to 1 quadrillion (1*10^15, or a peta prefix) brings us down to about an hour and a half. The next step is a quintillion (1*10^18 or a exa prefix), we’re now down to 4.5 seconds. OK, that's a little tangential, but it should underline the point of the state of EP nano-technology. It is economically, temporally, and physically possible to build whole goods using nano-fabrication. By the same token, it is also possible to use nano-tech to disassemble whole goods for recycling to produce new goods. I can't see a good reason that an insidious entity wouldn't attempt to manufacture something like this. Honestly, the big thing that I'm having trouble with are rogue industrial disassemblers. The little guys responsible for taking your garbage and breaking it down into useful components. Sure, they're going to be engineered with limitations, inhibitors, and whatnot, but through incident or intent these could be circumvented. This technology is essential for a nano-fabrication industry/economy. It's almost required. It's also one of the most dangerous things to have in a habitat. How much of the habitat infrastructure and construction is composed of the same material as the latest pop-tech fad? Cleaning up this stuff could be a royal pain, possibly an exercise in futility.

Again, that's more of a reason to slow down the overall pace of nano-fabrication. Actual disassembly by nanobots would be faster than assembly (no real reason to move, initially). There are a couple of propulsive methods that could spare some of the power. First, power on the outbound leg of the journey (from fabrication nozzle to the surface being created) could rely on external power. A magnetic coil or something could launch the nanobot towards the surface of the product. Carbon or silicon based nanobots should be able to withstand the stress of this type of impact. In fact, they might be able to re-generate some power from the impact. I know, at some point, hand-waving has to happen. I just want consistent hand-waving. If Grey Goo isn't possible because of "X", I don't want "X" to be violated during the practical nano-fabrication that pervades transhuman society. If Grey Goo is possible, how much of a threat is it? Industrial nuisance? Industrial disaster? X-Threat?

The EP world says nano-fabrication is not only practical, but widespread. How does it work?

There are plenty of reasons to say "It won't fly, Wilbur". The neat trick is figuring out how to make it fly, despite the naysayers. Even in fiction, there is invention.

Static DET5 wrote:

There are plenty of reasons to say "It won't fly, Wilbur". The neat trick is figuring out how to make it fly, despite the naysayers. Even in fiction, there is invention.

That's not a fair analogy. You're comparing me to people naysaying the Wright Brothers. Saying "You cannot fly a heavier than air vehicle" - this flying in the face of the *clear* evidence of heavier than air birds flying is one thing. Saying "This process is 40% efficient. So if you could get 100% efficiency, which isn't possible, it would only be 2.5x as good. So 20x as good just isn't in the cards" is another. You can have Santa Claus if you want, he's just not hard sci-fi though.

NewtonPulsifer wrote:

Static DET5 wrote:

There are plenty of reasons to say "It won't fly, Wilbur". The neat trick is figuring out how to make it fly, despite the naysayers. Even in fiction, there is invention.

That's not a fair analogy. You're comparing me to people naysaying the Wright Brothers. Saying "You cannot fly a heavier than air vehicle" - this flying in the face of the *clear* evidence of heavier than air birds flying is one thing. Saying "This process is 40% efficient. So if you could get 100% efficiency, which isn't possible, it would only be 2.5x as good. So 20x as good just isn't in the cards" is another. You can have Santa Claus if you want, he's just not hard sci-fi though.

Well, it's an Eclipse Phase board, with authors talking about Eclipse Phase (my apologies to our original author and host, who may have been thread-jacked. I really am interested in the industrial disaster part of this). In Eclipse Phase, one of the pillars holding the game world up is that of practical nano-fabrication and disassembly. So, how does that work, but not allow for insidious creeping crud? I love the game world and I love SF, so I'm not going to look at it and say "Well, none of it should work". I'm perfectly happy to look at it and say "Well, the current state of the technology allows for contained nano, but un-contained nano is pretty iffy". The EP core book is REALLY light on the various un-contained nano options. Sure, you can buy some nano that will slowly build you a gun, but there is no discussion about problems and issues with that build process. Is it not done because there are inevitable compromises, sometimes catastrophic ones? To me, that's cool. It also makes me think that somewhere, someone is working on perfecting the technique and is able to apply it to small, specific items. Maybe it is a matter of the massive scale of environmental interference. Sure, grey goo is technically possible in a clean room environment, BUT outside of a clean room it faces too many environmental factors to work reliably. The weapon becomes unpredictable, unstable, and essentially unusable because there are better ways to accomplish the same thing (Bioweapons have this trait). The destruction/disassembly is easy (If you allow for the technology in the first place. We have to. It's part of the game world). Turning around and synthesizing from that destruction is where it gets difficult. Without the ability to do that, your destructive element gets slowed at a cube pace as it expands to devour everything around it. Even with perfect, unfailing nano (Which we don't have), the process will eventually come to a relative standstill. It's still nasty stuff, and I think it would qualify as an industrial disaster. You'd have to bust your ass to make sure you got every bit of it, or you'd be seeing weird random failures throughout your hab.