« Climategate | Main | Imbeciles: the second round »

The myth of the starship

(NB: As starships do not in fact exist, no starships were harmed in the production of this essay. Also, this is just words. If they upset you, go lie down in a dark room for half an hour then drink a glass of water; you'll feel better.)

Actually, I tell a lie. There are five starships that we know of; Pioneer 10, Pioneer 11, Voyager 1, Voyager 2, and New Horizons. But they're a far cry from the gleaming interstellar transports of science fiction. New Horizons is the most recent of them. Launched in late 2006, it is the fastest human-launched vehicle so far. It raced past Lunar orbit within nine hours of take-off: nevertheless, it will take around 10 years to reach Pluto (its proximate target — for a three-hour flyby). It weighs around 478 Kg, and is currently travelling outwards from the sun at around 17km/sec — about fifty times as fast as a rifle bullet.

We are 4.37 light years, or 140 million light-seconds, from Alpha Centauri, give or take. One light second is 300,000 km; it takes New Horizons about five hours to travel one light second. So: in very roughly 30 million days, or on the order of 300,000 years (if it was going in the right direction, which it isn't), New Horizons could reach Alpha Centauri.

And that's the best we've done to date, admittedly without really trying ...

This is not an essay about whether we could do better if we tried. I've written about the problem of space colonization before. Rather, what intrigues me is the possibility that the entire conceptual framework of the starship is a dangerously misleading dead-end, and that what we need is a new framework for thinking about interstellar travel.

The very word "starship" is a concatenation of two other words — star, and ship. The first is pretty harmless; it merely defines the scale factor we're talking about, as opposed to interplanetary ship, or moon ship, or Atlantic-crossing steam ship. But the second word comes with a whole freightload of unwanted baggage, and I'm of a mind that serious futurists or SF writers might want to think about ditching it completely and looking for something new.

The astute reader will have spotted the link to the Apollo Program above. We have actually built and flown Moon ships ... but we didn't call them ships, and they didn't much look like one of these. There are several reasons why not. First and foremost are the scale factors involved, scale factors in both distance and (because you need to cross that distance) time, hence velocity (the change of distance over time). Kinetic energy is the killer; the kinetic energy of a moving body is proportional to the square of its velocity. Want to go faster? You need to throw bucketloads more energy at your vehicle, or make it much lighter. The Apollo spacecraft (or New Horizons) travel roughly three orders of magnitude faster than a regular sea-going ship, and they also had to haul along the reaction mass to throw out the back to get them up to speed (for they're all powered by rockets, relying on Newton's Third Law to make things balance out).

But there's a more subtle difference. We have a long tradition of nautical baggage. Seafaring ships of the great age of exploration were largely wooden, and — with the aid of their human crew — self-repairing; subject to the availability of raw materials, there wasn't much aboard a 16th or 17th century sailing ship that couldn't be made on board. Aside from carpentry, the inhabitants of even a relatively small port could make the necessities to keep a ship at sea on a voyage of years — a smithy, a pottery, a glass-blower, weavers of sailcloth and makers of hardtack. Shipbuilding was by no means easy (it was an economic activity born on the backs of the large numbers of peasant farmers and fisherfolk it took to provide the surplus to feed the workers in the shipyards) but it wasn't anything like the Apollo project, which sucked up the labour of a third of a million skilled engineers and technicians for a decade. The word ship therefore comes freighted with connotations of autonomy and sustainability that are inappropriate to space travel as we know it today. And one of the most perfidious, misleading, damning, unconscious associations of the word "ship" is the word "destination".

As I've said before, the trouble with going into space is that there's no "there" there when you get to the other end of your voyage. All you get is rocks, sunlight, and if you're lucky some slush (water optional: could equally well be methane or cyanide) to season your gravity wells. So if you want to do anything at the other end — anything beyond looking around real good — you need to bring the minimal seeds of the infrastructure with which to build and maintain your biosphere (if you're travelling with an entourage of meat puppets) or your mechanosphere (if you're going the Eric Drexler/Hans Moravec/downloading-or-AI route).

Which is why I was asking questions like this and this and this about a month ago. I was feeling my way towards this critical question: which is, "how simple can you make a minimal self-maintaining interstellar transport system"?

To a first approximation, the best answer I can come up with is "not very". We can probably make it mechanically simple, rugged, and lightweight if we can do mature machine-phase diamond-substrate nanotechnology, and if we can figure out how to do one of mind uploading or artificial general intelligence.

Note that I say mechanically simple — there's a monumental raft of complexity wrapped up in the idea of using a Starwisp for establishing interstellar transportation, but it's informational complexity rather than straightforward mechanical complexity. We may use such a system to sidestep the need for learning how to build self-sustaining biospheres and interstellar playpens for bored hominids, and how to equip a group of said hominids with the wherewithal to keep such a mobile playpen from degrading catastrophically, but we face the corresponding monumental challenge of solving the hard-AI problem and developing molecular manufacturing far beyond the flexibility and scope of today's nanotechnology applications.

Such an interstellar capability isn't going to look much like a "ship". It's going to look more like a DVD balanced on a microwave beam, or a can of beans hanging below a light sail energized by lasers powered by huge orbiting solar power stations. There won't be any biological agencies aboard: just AGIs or something equivalent ported out of a fleshbody's cranium. No hands, only nanotech assemblers. And after a voyage of decades or centuries it's going to have to stop — somehow braking at the other end — then spend more decades farming rocks, slush and sunlight to build ever-bigger physical structures until it can build the equipment with which to phone home.

If anything, it's going to resemble a seed pod for a different kind of life, and on arrival it's going to hatch and grow into a tree, or a forest, or a manufacturing-industrial complex. Finally, long after arrival, it might have sufficient resources to divert from homeostasis and growth to construct a biosphere, open communications with home, and prepare to download digitized colonists — if the whole uploading concept doesn't prove to be chimerical, and if there's something to be done with the serialized primate core-dumps at the other end.

Note that I'm fairly optimistic about mature diamond-phase nanotechnology (or some cognate thereof). The economic benefits of getting it are huge, and there are no obvious lacunae on the technology road map — unlike, say, fusion or manned interplanetary space travel. I'm less optimistic about mind uploading, because in neuroscience we are just about at the stage of beginning to figure out how ignorant we are. And I'm pessimistic about AGI, because I don't think we stand a hope in hell of working out how to design an artificial general intelligence until we know, at least in outline, what human general intelligence is. (And we don't.) But I suspect some combination of these technologies will show up sooner or later — barring resource-depletion crashes and/or habitable-biosphere-envelope departures on a planetary scale — and once you've got two out of the three, a starwisp-driven expansion starts to look feasible (if energetically expensive).

The alternative approach uses generation ships. That's what I was talking about earlier. Which doesn't sound much easier, if any, when you contemplate how much we don't know about environmental engineering and biology. It also begs some huge sociological — human — questions, and is unlikely to be planned as such. I expect it'd only happen in the wake of our development of stable, safe space habitats — itself a huge obstacle (Just look at the history of our first modular space station if you want to get an idea of the problems surrounding life in space) — and if the energy costs required to launch a starwisp are high (56Gw for about a month to punch a 1Kg payload up to 10% of lightspeed; a microwave lens 560Km in diameter to focus the beam), those of a generation ship are going to be astronomically higher.

... Which leads me to believe that we'd be more realistic if we just ditched the word "ship" entirely from discussions of interstellar travel.

What we need to contemplate are the requirements for of an interstellar transportation system.

Such a system needs to provide not only a mechanism for sending a self-replicating technosphere across interstellar distances; it needs to be able to produce a habitable space at the destination, and provide a return option (for data, if nothing else).

Waving a magic wand of the variety I discounted in my earlier space colonization essay, even if we postulate that mind uploading is both possible and the way forward, interstellar travel still won't be cheap. Direct communication via modulated laser looks feasible at extrasolar distances, on a reasonable power budget, given adequate pointing accuracy — but the study linked to above has a bit rate of only 0.5 kb/s. Given their 15w peak power output, they're talking about 0.03 joules/bit across 1000AU distance; or 1.9 joules/bit/light-year.

Even Hans Moravec's estimate of the computational complexity of the human brain (cited here only as a starting point for discussion of mind uploading: I think he's erring on the optimistic side by between three and six orders of magnitude) suggests you'd need many years to transmit a compressed uploaded mind at that bit rate! But we know we can transmit data much faster using lasers over optical fibre. If we can push the bit rate towards 1Tb/sec, transmitting a map of a brain with 1011 neurons and glial cells, each with on average 104 interactions with neighbouring cells, and, say, a spare 30 bits/connection (to summarize its properties), for a total of 1018 bits of data per upload, we can squirt that map at the stars in about ten days. We'll need a powerful laser unless we want the error rate to climb. Using the figure of 1.9 joules/bit/light-year, to send an uploaded mind to alpha centauri is going to cost on the order of 1019 joules over ten days: an unfeasibly large amount of energy. (Hint: we're talking the equivalent of a Hiroshima-sized nuke every second, for a million seconds).

However, that figure corresponds to the sort of bit rate we can envisage achieving more or less now. Improve the data-to-power ratio by, say, a millionfold and things begin to look feasible (if not exactly cheap). In particular, it's a lot cheaper than travelling in person (and anyway, we've got the honking great big solar power station/maser grid we built to launch the starwisp in the first place; might as well put it to use). Finally, if we can do the upload thing well enough to provide the brains to run our starwisp in the first place, we can probably make do without building a biosphere at the other end. In other words: frail fleshbodies need not apply.

(There's an alternative to shipping around uploads via laser that merits investigation: if we can do uploading, and if we can make memory diamond — which would seem to be a reasonable expectation of a mature machine-phase nanotechnology — then the 80g payload of the reference starwisp ought to be sufficient to carry about 2 x 1024 bits, which corresponds to 20,000 stored uploads per "passenger ship". This might well be energetically cheaper than using a laser to transmit uploads, giving us an unexpected long-haul corollary to Tanenbaum's law.)

So, to summarize: yes, I think human interstellar exploration (and yes, maybe even colonization) might be possible, after a fashion. But to get there, we're going to have to master at least two entire technological fields that don't yet exist, even before we start trying to blast compact disc sized machines up to relativistic velocities. And that's without considering the difficulty of how to cram an industrial infrastructure capable of building more of itself, of a machine capable of surviving in deep space — the equivalent of those 300,000 NASA technicians and engineers — into the aforementioned CD-sized machine ...

If we succeed in doing it, it's going to look nothing like the Starship Enterprise. Or even New Horizons. The whole reference frame we instinctively assume when we hear the word "ship" is just so wrong it's beyond wrong-ness: it's on a par with Baron Munchausen's lunar exploits as seen in light of the Apollo Program. We need a new handle for discussing and analyzing such a venture. And the sooner we consign the "-ship" suffix to the dustbin of failed ideas, the better.




Uploads are one thing; downloads are another. Reading the state of neurons, and their connections -- even destructively -- I can see doing. But building a brain with all the right connections set up, so that you can then somehow stuff the right data in them... that seems a lot harder to me. (But perhaps we won't be downloading to meat at that point -- just a body running a simulation of the brain.)


I think you are over egging the storage/transmission capacity somewhat. We know the brain has quite a bit of capacity to 'reboot' from little in the way of valid connection state data, plus data compression etc. Closer to 10^13 is my guess.


After I read Accelerando, I spent a while reading about diamond-based nanotechnology. And it turns out there's a huge technological barrier.

Since Drexler wrote Nanosystems back in 1994, there's been a lot of research into surface physics. And it turns out that Drexler's rough sketch of 2D diamond surfaces was far too simple. As I understand it, there are all kinds of nasty quantum-mechanical deformations that occur as the diamond tries to reach a minimal energy state.

For Drexler's basic design to work, he needs to be able to deposit a carbon atom with very high accuracy: Something like 1 error in 10^15 atoms. But with several months of compute time on a big cluster, the best error rate (as of a few years ago) was something like 4 errors for every 5 atoms placed.[1] Every simulation ran up against the limits of computational chemistry. At the time, most of the people who studied the surface physics of diamonds seemed to be very, very pessimistic about machine-phase nanotechnology.

[1] I'll spend a few minutes digging for the papers and see what I can find.


"Heavier-than-air flying machines are impossible" -- Lord Kelvin.


Ah, here we go: Theoretical Analysis of a Carbon-Carbon Dimer Placement Tool for Diamond Mechanosynthesis (PDF). Poking through Google Scholar, I see some more followup work by Freitas and Merkle. But they still appear to be the only two serious researchers currently investigating diamond mechanosynthesis, AFAICT.

Richard Jones, the author of Soft Machines, suggested that researchers would be better off pursuing Brownian motion, self-assembly, and all the other tricks of conventional biology. Using these techniques, you might eventually be able to construct some kind of weird, synthetic ecosystem in a can. But I'm not sure that would help you much if you could only drop 1kg of cargo on a distant planet.

Is there a reason why you've ruled out Orion drives?


Charlie's new unit of energy: the "bucketload".


Terrific work Charlie.

I couldn't find anything to disagree with, except your statement that we need to understand human intelligence before creating artificial minds.

Why do you think we need to understand the human mind to create a non-human mind? Evolved systems are very hard to understand compared to designed systems.

For example, it was pretty much impossible to figure out how old freight train pneumatic braking systems worked, but they did work. They evolved. On the other hand, it's easy to understand how modern electronic breaking systems work. They were designed.

Bird flight is very hard to fully understand. Jets are simple.

There's no reason an AI has to think much like we do. So I don't share your optimism about AGIs being hard to do. I fear we'll see them before I croak, unless I manage that trick before 2040.


What about sending a small quantity of quantum entangled mass to transmit the "brain uploads" once the starwisp reached its destination? Sure, its an unproven, undeveloped technology that may or may not work but so is a lot of the technology that you propose in your posting. If it could be made to work, you would have an instantaneous, low power receiver to transmit the brain scans of your travelers into their new metallic bodies.


This makes me think two things;

1. It's a great resolution to the Fermi Paradox - it's just too expensive (energetically as much as financially/whatever aliens use) to do it often, and

2. All sorts of good story possibilities about what happens if such a starwhisp reaches Earth and starts to 'upgrade' our habitat.


Which science fiction novel was it where scientists intercepted an interstellar signal that turned out to be instructions for building a machine, which in turn grew an organism (a human-shaped alien, I believe)?

This was posited as the most efficient way to spread to other planets — make the locals reconstruct you at the other end.


I'm pretty much with you about shipping "uploads." What those uploads should be is another question.

We're assuming a couple of things

The important question is what uploading and downloading is. Mapping someone's brain is a waste of time, unless you have the way to assemble that brain in a living state. That's akin to assembling a running car with 20,000 km wear. It's silly, and possibly stupid. After all, the downloads will be incarnated in a different environment than they were uploaded in.

The minimum you need to ship is a) a lifelog, and b) an expert system simulation that responds just like you do under all known situations. If you want to create a "download" at the other end, clone a baby with a computer implant, and give the kid access to the lifelog and the simulation. Kids are "designed" to adapt to the environments they find themselves in, and giving them conscious access to their previous lives will allow them to learn from their old incarnations, plus expert advice from that old person, will, in many ways, allow them to become those uploads.

Thing is, if a person colonizes a new planet, only some of their old experiences are going to be relevant, and they're going to change anyway. This version of a download is only slightly more extreme.

I suspect it would also be a lot more useful to send a simulation than a brain map anyway. It lets you standardize a lot of things in the design.

As for the starwisp, I'd be scared by a culture that could hit a target moving near the speed of light from light years away. That's the kind of technology that would allow me to put a laser dot on the third left upper proximal eyelash of an astronaut on Mars, if I wanted to, accounting for atmospheric distortion. Pretty hard to do. I don't think that starwisps or sails are possible over interstellar distances, just because of the problems aiming the beam and keeping the beam-space clear behind the ship. The Orions Arm beamriders might work (they're proposing self-steering "smart matter" shot from linear accelerators, instead of light beams). Something like antimatter or fusion rockets might work too, especially if you combined it with a magsail brake to slow the ship down to its destination, rather than firing the engines to slow down.


@ 6 dolphins use 1/7 the power moving through water than is needed to drag their form, I expect that birds are much the same. Evolved things seem to work quite well, compared to made.

@ 8 well, that is one way to look at David Gerold's Chthor series.


Over a decade ago I idly mused about an interstellar "empire" that didn't use ships because, as you point out, the energy costs were absurd... but I cheated there, assuming that syphoning about 10% of the Sun's output (in visible light) could be used to create the exotic matter near-handwavium stuff needed for those Einstein-baffling artificial wormholes that other, more mathematically-able folks were idly musing about at that time. "Cheap" (relatively speaking) starwisp-like probes would scout out likely targets and those reporting back promising results would be assigned priorities for one of these gates... which'd be shipped unmanned and notably STL, taking decades to make and centuries to arrive.

Not exactly Captain Kirk's cup-o-tea, I guess, but it was interesting to muse on what kind of society could support such an enterprise over so long. (In the end I again cheated somewhat and had a "smart" AI, whether it was self-aware wasn't something I bothered considering, in charge doggedly pursuing its assigned mission regardless about what the meat wanted to do. And with those energies in play, the meat was too nervous to risk joggling the AI's elbow into an accident or making it miffed and resulting in an "accident".)

But yeah, the concept of an actual "star ship" would require remarkably new physics to make plausible. Alas.

-- Steve


SpudTater@9, that's A for Andromeda, by Fred Hoyle and John Elliot.



Why don't we just take our habitat with us? Perhaps some sort of self-sustaining ring world? Build it in our neighborhood, give a little push, and settle in for the long haul.


That link to Moravec's page is 404ing.


Drexler considered that question of interstellar travel given advanced nanotech: http://e-drexler.com/d/06/00/EOC/EOC_References.html#0115

His sketch didn't look anything like a ship, either, but he didn't think it'd take decades to build structures at the destination.


The Moravec link doesn't work for me, either.

What capacity to reboot, Ian@2? You get a working human if lose alot of your connection state, BUT it's hardly the same human, is it? He'll have memory and skill problems when trying to continue his life post-trauma. Even tiny amounts of ongoing neurodegeneration pose serious challenges to people. No, Ian, the realm of compression will be trying to keep it down to a mere 10^17th, as I see it (and possibly failing).


I think starwisps would lend themselves more to r than K reproduction strategies.

It probably doesn't matter much how long a starwisp takes to get where it's going; something that will last in space for 100 years is probably good for 1000, at least 10% of the time. Probably better. Since the energy budget is going to be almost exclusively spent on propulsion, why not send out 10 (or 100, or 1000) slow starwisps rather than one fast?


Tell me again why any human is going to another solar system?

If we had that level of tech (AGI, nano), we'd easily be able to colonize our system -- every rock, every comet, every planet. Are we going because we need the space? Not likely.

Are we going for the resources? Can't get anything you find back, and nothing out there is worth the expense.

And unlike you, I don't expect the biological human race to continue very long after the point where we have uploads or AGI. That will be the cutting edge and the bots will be making all the decisions.

Why would bots send meat into space?


Personally I dont have a problem with the ship idea, its the star part that bothers me. The technological hurdle in the creation of ships for space is tangible. But getting them to go to other stars is a stretch.

Theres a article on NewScentist atm suggesting using Black holes or Dark matter as a means of propulsion.

As an idea this sits uncomfortably. Dark matter is still a theoretical dream. And so are Black holes. The usage of the word black or dark in these theoretical suppositions is to express the unknown. Human pundit for catchy names tends to give the phenomena a mental reality. I guess the same applies to the word ship. (Darkship?) :)

But these ideas dont come from nothing. Observation is key here. We have data to which we ascribe a theory and give it a name. The objects in question however continue to be subjective in terms of human comprehension. We interpret relative to our comprehension. i.e. before the discovery of the atom the best theories for what powers the sun was coal.

Consider astronomy. One subjective realization I have come to is that our observations are interpretations. Telescopes augment our perception. ie, we see in near and far Infra-red way beyond our biological capabilities. The pretty images that we are presented with are a visible light interpretation of the data.

OK, my point is, we are already bringing the universe to us via telescopes. In a way we are already traveling through space by proxy. The cognitive experience and the tangible physical presence 10 light years away are one and the same. Its just an interaction with the data.

If you are an uploaded entity in a Starwhisp, how is the subjective experiance any diffrent to interacting with collected data here on earth?

I make this point because our linear definition of time breaks down when faced with the prospect of interstellar travel. Whichever way you look at it, interacting with the data of the cosmos incurs a time debt.

What we are ultimately dreaming about in the concept of a starship is a kind of cheating. A means to not be bound by the linear speed of light and time. Timelords?

There is much (theoretical darkness) still in the universe that needs to be bought into the light, (so to speak). There is hope yet in the theoretical dreaming of the human race that may bring this goal closer. Take black holes. Our interpreted view of these things is actually a bit evil. In our musing they are the antithesis of light and life. But these theoretical constructs are the Hubs of galaxies that create energy, order and life (us by example).

I guess that the more we try to understand a concept by study that less intangible it gets. There is still a lot to learn that may change how we conceptualize the idea of space travel. Your reference to understanding (what we are) is a valid one. As much as the concept of a transcendental state (aka Uploading) is a catchy one atm, its one that also sits uncomfortably. One gripe I have always had with fiction is that it sees our biology as a inferior to the technology we create. My view is that our biology is a magnitude higher technology, just because its tuned to our biosphere is a moot point.



@12: Anton, did you discuss any of your ideas with the Orion's Arm group? This sounds really, really familiar.

@18: It is a bit unclear why humans would want to go to another solar system, but our upload lords and masters might think that keeping humans alive in another solar system qualifies as a truly nifty and cool status symbol. How does the song go? "When the uploads come, we'll make great pets?"


Human brains should compress fairly well, especially if you're willing to accept a bit of loss. With standardized education, rigid social norms enforced by peer pressure, and mass media, we're not as different from each other as we'd like to think. Suppose we scan a few thousand or a few million minds, identify the common or generic experiences, and pre-load the starwisp with a library of them. As it travels to its destination, we'd keep transmitting updates, accommodating cultural drift.

When it comes time to transmit your mind to a distant star, you'd be able to decide if you really wanted to send 23 terrabytes of your memories of second-year differential calculus class, or just transmit "install maths_library.h".


@20: Nope, I'm unfamiliar with the Orion Arm project... any parallels would most likely come from me cribbing from the same sources they did. (Of course, they actually did something other than muse on it, so more power to them.)

Now that I've browsed them a bit, they took the idea in a rather different direction than I did. I'm afraid that my civilization turned out a lot more "water empire" like than the one they propose, and maybe a tad duller. (Much less cyberpunkish, too; I had my "singularity" event fail rather nastily to give my limited technological imagination some breathing room.) Mind you, I got to squeeze some ideas for sunjamming Aubrey-Maturin knockoffs out of my setting so we'll call it a draw. *grin*

-- Steve


@18: That's exactly the point that Charlie is making--it's going to be bots, not meat. Can't be anything else. And for why--well, who knows? Maybe it's something like machine Puritans (the kind that ended up in Massachusetts). Fleeing a thought environment they feel is hostile, and the ruling overlords thanking them for it. Of course, these guys have to be really wanting to get out of town for this...Thinking of a sequel for Saturn's Children, Charlie?

@OP: Well, 10^19 J over 10 days is not so bad if you have substantial solar radiation gathering. It's about one ten-millionth of one second of solar radiation output. And you're proposing giant high-powered solar-fulled lasers for getting the things to go fast, so...

@17: Yes, that makes sense. Especially given Charlie's point that these things resemble seeds more than ships, and what kind of strategy do trees use with their seeds? I'm not sure under this scenario that there will be any incentive for others to do the old colonize the target first thing, as there are a million other stars, and it doesn't make a lot of difference how long it takes to get there.



I do not believe general intelligence can be artificial, far less "designed".

What is intelligence? It's Maxwell's demon, a self-sustaining (and sometimes self-conscious) entropy-retarding structure. Actually a local entropy inversion within a flow state system.

Is it possible in silicon? Definitely yes.

How would natural general intelligence evolve in a silicon environment?
Easy. It needs a stable entropy differential as food source, external turbulence (stochastic and non-normally distributed in frequency and extent) inducing survival pressure, heavy competition and mechanisms for change and transmission of genes/memes or other forms of information.
All this exists. It is there. In the financial markets you have all ingredients of evolution present.

I truly believe that 100 years from now (if mankind survives so long) AI designers will be mocked not only by humans, but also by naturally evolved sentient computer-based life forms.


You'd only need to send one complete human mind. Everything after that would just be a "diff," and maybe an MD sum for the diff, so bandwidth isn't a huge problem... And if you could get some relays into place the power requirements wouldn't be too large. (Your error rate would decrease rapidly once the signal left solar space.)

The relays might be your other starwisps - if you can solve the power problem - maybe you could modulate your propulsion beam to provide signal? Just some late night thoughts...


Charlie@OP: From the destination world's perspective, what you're proposing is a 1 kg gray goo bomb. Long before you have the ability to seed gray goo on a distant world, you'll have the ability to release it at home. That won't require the propulsion system, the energy, or the miniaturization. For a long time while the starwisp is in transit, you'll still have the capability to destroy the home world with goo. So the starwisp designers can assume that the homeworld will probably be gone before the starwisp arrives.

truth is life@23: There is incentive to compete for nearby worlds. If several groups (nations, species, operating systems, whatever) are trying to achieve dominance in sheer numbers, growth rate matters, and stealing a planet 100 years away is faster than shooting for an unclaimed planet 500 years away.


J Boss.

A society made up of AI bankers?

If that made it into Doctor Who, would it be called The Wunch.


John @7: I don't think we need to understand how the human mind works before we can design AGI, but having a clue what intelligence is would help a little. I'm not convinced that we know what we're talking about when we discuss intelligence, or consciousness, but we think we do ... and it taints all our discussions.

Minor digression: I had a real throw-book-at-wall moment years ago with Roger Penrose's second attempt at proving AI impossible, "Shadows of the Mind", when he glibly asserted early on that we don't need to understand something to disprove the possibility of emulating it mechanically -- in his case, consciousness. Oh, really? In any other field that statement would have gotten him roundly mocked, but none of the reviewers seemed to have picked up on it ... it's symptomatic of the confusion surrounding the subject that fifty years after the AI project got under way, we still don't have a definition of the target that makes sense.

h @11: you don't aim the beam at the starwisp across interstellar distance; rather, you rely on the starwisp to ride the beam and keep itself centered. (After you shut down the boost-phase beam, you keep a guidance beam running at low -- but sufficient to remain detectable at interstellar distances -- power until it's time to power it back up for the decelleration phase.) Stuff I didn't mention: the idea of using a two-stage light sail, where the second stage decellerates the payload at the far end using light reflected by the first stage (which, by conservation of momentum, keeps on accelerating past the destination). Otherwise, I think you're pretty much right.

mgoodfel @20: exploration. Starwisps probably aren't very expensive, if the enabling technologies are there and we look at them from the PoV of a post-AGI culture. (Mind you, the political aspects are there: which of your neighbours would you trust with a 56Gw maser that can be colimated to a hundred metre beam at a hundred trillion kilometres range? We are talking Dr Evil grade death rays here ...)

kbob @28: "grey goo" implies uncontrolled, undirected replication; it's not considered likely. (We have an existence proof for the non-existence of grey goo: ourselves. We live in a wet-phase nanosystem, namely the DNA/RNA/protein world. It's had the thick end of four gigayears to evolve grey goo -- if wet-phase grey goo, the Eater of Life, was possible: but all we've got are some rather agressive bacterial populations. No hegemonizing slimes that break down everything in their path! So I'm optimistic on "grey goo" being difficult to create. An intelligently-directed self-bootstrapping industrial complex is another matter, but I'd expect any such thing to have so many critical paths that it's not going to be much of a risk for runaway growth, at least not in the early stages.


JBoss @26: What makes you think that the goal of an "AI designer" is to build general intelligences? The vast majority of modern AI research is about solving specific problems (or class of problems) using methods inspired by natural intelligences.

And the use of evolutionary pressure (in either natural or artificial environments) to solve problems is old hat by now.

PeteY @14: That's the one, thanks. Didn't realise it was a novelisation.


Kris@7: Sadly, you can't actually use entanglement to transmit information. You just get the same *random* state at both ends.

It's possible to transmit an arbitrary amount of information with a fixed amount of energy: you encode the information by the time or direction the energy is sent from, or the frequency it is sent using. Doing so requires an increasing amount of time, space, or bandwidth, though.


The seed idea seems to me to be locked into a two pronged dead end street.

One the one hand, we send "humans" out there, in their squishy form. However, a "human" is not a brainscan and a pack of genes, it's an entire symbiotic ecosystem, both within our bodies and outside. We'd need the ability to recreate humans, and a myriad of other creatures, in a roughly goldilocksy just-right stew of organisms to keep the squishies alive (miss the wrong microbe, and the whole thing collapses). Putting aside the increased data load for our interstellar package, it looks to me like we would need to pack an entire terraforming seed into a wisp, or have our travellers confined to an extremely unstable lonely bubble of life in an extremely hostile enviroment.

On the other hand, if we are sending mechanicals by way of uploaded consciousness and nano-replicators, it kinda seems to me that there is really no good reason to actually do interstellar travel as such other than an eganesque high-risk, low-payoff idealistic "pure exploration" drive to *be there* in our non-fleshy incarnation and do... what? Have a look around? Or are we establishing a... mechanical colony? In which case all that is needed is some form of computronium on the other side, which means that we're making a VR enviroment that is *there*, and not *here*... kinda pointless, considering the expense of doing it.


Sebastian: you're right, but there's an avenue you haven't explored: exploration.

a) Axiomatic: if we want to actually send a probe to examine a neighbouring solar system, close up, then it needs to be autonomous, for values of AGI or uploaded-human-mind-in-control. Anything less won't be capable of doing the job because the one thing we've learned so far from fifty years of planetary science is that solar systems are complex and surprising places.

b) It follows that it would be useful for such an autonomous probe to be capable of (i) maneuvering to examine different types of bodies, (ii) building sub-probes to descend to the surface of different bodies, and (iii) communicating its findings home.

c) It is hard to imagine how to cram an interstellar communications system and/or surface probes into a 1Kg payload. So it's going to have to construct them once it's out there.

d) Even across interplanetary distances, control/command of remote probes is difficult (as witness the fun NASA are having with one of their Mars rovers right now). So it'd help if the locally-built specialist probes were autonomous, too.

e) At this point, you've got the manufacturing capability to build a mechanical colony, purely to support the exploration role. So why not keep building out and use it for other purposes? From the PoV of the people back home, it's already a sunk cost. And there's no gain in actively shutting down a self-extending probe infrastructure once its initial exploration work is done.


@9 pt 2
Already been done: "Kirinya" by Ian McDonald.

The Genisis Project, by Donalld Moffitt - except it is the Humans who arrive at the other end ...
And "A for Andromeda" Hoyle and .... erm ....

What IS clear is that we need handwavium of SOME sort.
Preferably a physics breakthrough or alteration-of-viewpoint.
An EFFICENT "reactionaless" drive would be good, for instance, allowing us to get bodies up to really singificant fractions of c ...
OR a "wormhole" option, allowing pop/skipping across space in big jumps.
Though that idea led to one of the best-loved SF stories of all time

Beware of dragons.


@35: 10 is Contact, Sagan, isn't it?


At what point does a society of uploads have enough population pressure to require colonization anyway? It's always going to be inward facing, culminating in a matrioshka brain style maximizing of the local resources, but how do we reach that point? Are the virtual people in the virtual world having virtual babies at exponential rates?

It's probably my anthropocentrism talking but colonization by starwisp uploads seems pointless to me.


Nestor: insufficient data.

We have no idea whether an upload society would even retain the human traits of consciousness and individuality. Might it consist of lots of monadic individuals with low bit-rate communication channels between them, or would the availability of broadband connections between minds (functional telepathy) swamp individuality? Would they archive past states, or prefer rolling amnesia in order to retain a sense of novelty? (Noting in passing that our addiction to novelty seems to be a side-effect of certain dopaminergic pathways in our meatbrains ...)

Your guess is as good as mine.


Charlie@34: all true, but then we are not talking about interstellar travel or even colonization, more of an interstellar mechanoid infection/infestation by autonomous agents that have little to do with us (I'm not very good at being a space telescope/spectroscope/communications facility/etc. and doubt that you are, either).

Nestor@37: once it runs out of storage capacities? Even at nonexponential rates, you hit the limits of storage and/or computing capacities sooner or later. Rudy Rucker had fun with that concept in Postsingular.


Spud @31:

I am not sure, if you fully grasped what intelligence is.
Goal oriented intelligent systems are easy. You have feedback. You can add in evolution, which is easy again, because goal-oriented intelligence by definition has a benchmark. Evolutionary pressure is directed and one-dimensional.

General intelligence is goal-independent. Different universe. What is general intelligence exactly and can you measure it? Or which kind of evolutionary pressure enhances something by definition undefined?

This is unsolvable if you don't change the approach.

So leave the easy feedback approach, conceptionally a little Turing thingy that can do all kind of tricks. Even a Turing machine cannot do something undefined. It wouldn't know how to stop.

General intelligence involves a higher level of complexity. A Turing machine defining it's own goals and deciding on its own, when it wants to stop. You think you can build that? Good luck.

The alternative approach is to start from a food source (entropy differential) and hurl problems at it. Undefined and unforeseeable problems. Having access to specialised expert systems it can cannibalise or integrate as subsystems helps. But in the end, it's only general intelligence, if it's not goal oriented. And if it's not goal oriented (besides the obvious goal of survival), you can't control its intentions.

General intelligence is an infestation. You wouldn't even want to design/construct it under normal circumstances. What for?

But given the food source, the building blocks and time (processor time in this case), it's still inevitable.


I know you are against the whole space opera concept, and I repsect that, but trying to literally take what we know today to build something in the near term to explore beyond our solar system is creating its own problems. How much more energy can we create in a burst form now than we could 100 years ago? 500?
Graphene composites may be the answer to faster computing and I can't even begin to guess at what forms of engery we will be developing in the next 200 years. Nanotech is in its infancy and industrial nanotech hasn't even been truly envisioned yet, much less realized. There will be new types of materials that are much stronger and lighter and new forms of energy that we have trouble imagining now. I understand that the entire of point of this is to be able to imagine it, but I think you are hamstinging yourself too much. Call it handwavium, but I guess I'm hopeful that we will develop better ways to make stuff and power stuff. Assuming you can do that it still brings you back to if you do decide to carry the meat, how can you do it safely from a physical and psychological standpoint.


I'd like to point out another piece of baggage implicit in calling it a "starship": A command/leadership structure that parallels (for English-speakers, anyway) that of the Napoleonic-era Royal and US Navies. One of the almost-unchallenged memes of space opera is that the decision-making structure will look like Horatio Hornblower's. Leaving aside interservice rivalries (hey, I don't have to acknowledge the technological skills of squid!), there's a simple and egregious problem with a "naval" command structure: It implicitly assumes that the decisionmaker has fairly detailed understanding of all of the technology and knowledge relevant to the decision.

Ain't gonna happen in space.

Ain't gonna happen in space when contacting Houston for guidance requires hours/days/years for messages.

So even calling it a "ship" of any kind brings along baggage. And since baggage is heavy...


@43: I think you have this backwards. The autonomy of a Napoleonic-era captain operating far from home is exactly what's needed for NAFAL space exploration beyond the solar system, for exactly the same reason that the Crown gave them that much autonomy in the first place -- communication was so slow. And that meant "months" back then, where we're talking "centuries".


DD-B: a military command model sets you up for a single point of cognitive failure, though. I think CEP's argument is that the complexity of the systems we're discussing is such that no single human decision-maker can have a grasp of every aspect of the technologies in play.


Eric@3:IIRC, diamond is already thermodynamically unstable. Left alone, it will tend to 'decay' - if that's the correct use of the word in this context - into graphite.

"A for Andromeda" is of course the first and by far the best riff on the subject[1]. One point the author seemed to be making is that uploads aren't the way to go. Suppose that it takes 10^18 bits to correctly specify an uploaded mind, and suppose that the transmission is interrupted so that only 9.9*10^17 bits are received correctly. If this is a remotely human intelligence, what you're going to get will be in all probability just random hash.

So what is called for in this instance will be uniquely constructed 'Agents' that can be transmitted module by self-contained module. Some sort of initial block of information that specifies the executive control layout would be the first and most important component transmitted, followed by successive modules that would correspond to memory, different modes of information processing, etc. With something constructed in this fashion, half or more than half of the signal can be lost and you still get a useful mind on the other side. Iow, the intelligence that you'll actually be interacting with at the far end won't be even remotely human(though it will represent them, just as Hoyle's embodied intelligence carried out the will of an extraterrestrial civilization.)

[1]If anyone has a better candidate for this type of story, I'd certainly like to hear about it.


The simple solution is:
Step 1: sling a few tonnes of deinococcus radiodurans at some promising planets.
Step 2: wait for intelligent life to evolve at the other end.

You get up to the tonne level because when you're allowing hundreds of millions of years for the evolution of multicellular organisms, a few million for transit is peanuts.

Okay, it won't give us a humaniform star-spanning culture - at least, not right away. But given that we seem likely to someday sterilise this dirtball in a moment of absent-minded arrogance, I wonder if we almost have the obligation to try something like this. We probably have the tech right now.



A for Andromeda - by Fred Hoyle and John Elliot is the classic.
Andromeda Breakthrough - is the followup.
Ossian's Ride - by Fred Hoyle is what would happen if the entity was allowed to function quietly in society, i.e., technology transfers.

Here is the original BBC TV series on Youtube (harvest it while you can.)

A for Andromeda

Face of the Tiger




Then there is _Contact_ by Carl Sagan, to build a device rather than a person. And the _Species_ movie series where the entity was sent to wipe out competing worlds. The the movie _Virus_ where a signal was intercepted and AI life starts destroying the meat it finds infesting the ship.

BTW, The various starship threads reminds me of the armchair Victorians discussing the technology of the 21st century. The reality only occurred when real engineers, through trial and error, made the technology of today happen.

Put another way, we are lung fish, just now breathing air and struggling to get to the next water hole, with no clue about the next million years of evolution leading to various armchair Victorians arguing over starflight. I wonder if that lung fish would have made the effort to colonize the land if it knew that you guys would be the result.

Rather than remain armchair Victorians, we need to get out into space and find out what is real. Everything else is hand waving. Go beyond the Mundane SF movement and into reality. The real Universe is far more interesting than you realize. Consider these simple facts:

There was no Big Bang.

There are no Black holes.

There are no Neutron Stars.

The Sun is not an exploding hydrogen bomb.

The Universe is Electric.


charlie@44 - so you think that the captain of a nuclear carrier knows everything about his ship? (And after your "memoirs" of tech life too!) I doubt it very much. There is a hierarchy of knowledge in the command structure that allows each command level to have the appropriate knowledge of the sub-levels. I think this is all part of Ashby's Law of requisite variety. Therefore I think David@43 is correct.


As Greg Tingey@35 points out, the assumption is that mass cannot in any way be circumvented in our technologies. Change that assumption, and we could have Starship Enterprise like ships, probably even FTL. It might seem to violate the laws of physics today (conservation of energy), but must it be so necessarily? Suppose we find that we can remove the mass from matter and reacquire it at our destination, then we have not violated energy conservation when we move our massless (inertialess) ship to its destination.

Of course it is magic today, but will it necessarily be so by the time you even launch your nanotech starwisp?

On a more prosaic note, suppose we have black hole drives like that posited by Clarke in "Imperial Earth" (c.f. recent New Scientist), wouldn't the sheer mass of the black hole mean that you might as well build a big ship around it as the mass increment is small compared to the mass in the hole?


Concerning starship autonomy. The assumption that the starship cannot make the parts it needs is also flawed. We may well have "fabricators" that can manufacture any part from an information blueprint. In addition, if the most of the ship's fabric was similar to biology, it could probably self repair using basic materials available anywhere. As a result, I don't see that the starship must necessarily be too complex to maintain (which was an issue raised with the generation ship too). By the end of this century we may know an awful lot about designing self organizing systems on biological principles that can create remarkably complex artifacts that may only need advanced "gardening skills" to maintain.


Why are we ignoring Sleeper Ships again? I think freezing people will be much easier than uploading them (which, I think, is utterly impossible). And it will be easier than Generation Ships, because you avoid all social problems, and don`t have to sustain a biosphere in space.


"Shipbuilding was by no means easy (it was an economic activity born on the backs of the large numbers of peasant farmers and fisherfolk it took to provide the surplus to feed the workers in the shipyards) but it wasn't anything like the Apollo project, which sucked up the labour of a third of a million skilled engineers and technicians for a decade. The word ship therefore comes freighted with connotations of autonomy and sustainability that are inappropriate to space travel as we know it today."

I don't buy that argument at all. Ships today, e.g. nuclear carriers are hugely expensive to create with regards to materials and manpower, only partly autonomous yet we call it a ship. It has no destination either, just sea trips.
Likewise, aircraft are very similar, but we don't call them airships, although we did use the term for lighter than air dirigibles. Aircraft have small crews, but the same sort of command structure as ships. Of course super freighters are also still called ships even though they also have very limited crews and command structure.

Therefore, even though ships do have cultural baggage attached to them, the size of that baggage depends on the observer, not on the artifact itself. Had the Apollo spacecraft (command module, LM and service module) been much larger, I would have been quite happy with the term spaceship or moonship.



Alex, what you describe is one step away from a self-replicating spacecraft. And those beasts can and WILL undergo natural selection, so you will soon (relatively) have the entire universe filled to the brim with spaceships shooting each other for resources. Which, undoublty, will be fun to watch. From another universe.


If the transport was a storage device for human minds and was also a running a simulation of a star ship for those minds (ala 2nd Life/the Matrix), wouldn't it be fair to call the transport a starship - as that it what it appears to be to the occupants?

At the "destination", suppose there was some sort of biosphere/technosphere development, then couldn't the human minds still interact with the external reality via machines, robotic bodies, etc even if only newly gestated biological humans could actually inhabit the "real" world? All very P K Dick, I know, but why not?


Anatoly@50 - you are correct, but sterile biology exists even on earth where life only continues through replication. It also takes a lot of changes to overcome it, so I would say that it is extremely unlikely to happen to the starship fleet you have. About as likely as particular mule will have of siring/bearing offspring.


AAAAAND ... the dingbat prize for crank of the thread is won by allynh @47!

(Reminds me of an old Thomas Dolby concert rap, winding his way up to ... "the Earth ... is ... flat!" Except he was consciously doing it to amuse, unlike allynh, who really seems to have a bad case of Engineer Syndrome.)

Anatoly @51: because we've been there already. (See also the minimum ecosystem thread.)


Charlie @56

We`ve been at all kinds of starships before, I think. It`s just that you prefer mind uploading and I hybernation. And we will probably see neither in our lifetimes, but I still think hybernation is closer. *sticks a tongue*

Question about uploading, though - are you proposing to COPY minds or to actually TRANSFER them? And if it is the latter, how will you prove it is the same individual on the receiving end?


Alex @55:

Actually, once we have the technology, all it takes is one bored human launching one self-replicating ship. You know, for shit and giggles. 8-)


Anatoly@58 - when you have that mule replicating, let me know :) (actually I'll be able to read about it in the media, it would be a major story).


Alex @59:

Dude, check your wikipedia. 8-)


Anatoly@60 - I admit I hadn't heard of these mule cases. Even here, they are exceedingly rare - female mules mating with horses or donkeys and producing a foal. So our bio starship will have to be engineered carefully to ensure that they cannot replicate out of control.


Note that you don't have to be perfectly error-free in the mind-data stream-- the biological information store you're emulating is certainly -not- error-free. It's more important that the error-detection and correction scheme is smart and robust.

In fact, I wouldn't be surprised if brains have a randomness generator in them-- so, e.g., the old RAND books of a million random digits have a mildly apologetic note at the start, about how there was some proofreading of the random data, but not a whole lot. Same with mind/brain, I'd guess, so that the encoding requirements are not at the bit-level. But getting the encoding right matters a lot, and correcting the things that matter also matters a lot.


23 - Resonant: That's a pretty crazy idea, actually. What if, in an attempt to digitize a human brain using an imperfect algorithm (e.g. it only picks up general features or has inadequate density scan, or an error rate, or something), an average of many, many human brain scans was used. If you sent that over to another star system, imagine that human trying to figure out who or what it was or what it did? Or what it was supposed to do?



How about calling it "vessel" instead of ship? In vacuum it's all the more appropriate.

Also, doesn't having a 1kg probe, using an 80g diamond as a memory to build itself the infrastructure to phone home contradict your statement that a star vessel is a mechanically too complex affair to allow the crew to build everything inside of it?

The biosphere is a whole different matter.

I'm not sure if a star whisp is truly any more (not necessarily less) realistic than a large star vessel.


"...map of a brain with 10^11 neurons and glial cells, each with on average 10^4 interactions with neighbouring cells, and, say, a spare 30 bits/connection (to summarize its properties), for a total of 10^18 bits of data..."

This may be orders of magnitude too small or too large. Too small if it turned out that neuromes are very complex in terms of their connectionist properties dues to their receptors and protein status, or too large if it turns out that there is redundancy in the network to ensure robustness in a probablistic connectionist system.

However this does not impact the overall conclusion that it may be better to transport bits physically than transmit them.



The most likely outcomes for early-stage upload society are Vinge-style ubiquitous law enforcement or a Stepford Wives scenario. If uploading happens, there will be irresistible pressure to monitor and control what uploads are thinking. For the good of the shareholders and the safety of the children[1], of course. It's only a very short walk down that path away from innovation and exploration and towards an enforced stasis based on rentier pseudo-captialism.

Don't assume uploads will be thinking much of anything beyond thoughts designed to make the owners of the upload substrate richer.

[1] Or the rights holders, or the state, or whatever other shibboleth is most appropriate for the place and time.


Oh, and instead of a 'ship'... How about a Container?


Considering wishful concepts like mind uploads now is like considering a moon landing prior to the invention of gunpowder - it's a line of thinking that inhabits the realm of religious mythology. In the 'golden age' of science fiction, did fans talk about spaceships as if they would be real one day? Maybe we should ask Fred Pohl.


Speaking of concerts (your reference to Thomas Dolby), how was the Gary Numan gig?
(I know that's a pretty tenuous way to find an excuse to go OT, but I was quite bummed that you referenced it in a post that had comments off because the rest of it was so tempting to trolls; I was most interested in that throw-away mention...)


Regarding uploads, is everyone here up to date on the latest news? They're up to "larger than cat cortex" simulations


Granted this is not an actual model of any specific brain so it's not an upload in any way but it seems to be technology heading in that direction. Last thing I hear about this they had a single simulated cortical collumn equivalent to part of a small mouse brain. They also claim it will scale with Moore's law.

Fascinating stuff.


@68: You do know there's such a thing as fiction, and that the proprietor of this blog makes a living writing it?


@@3 Drexler's actually blogged about that. Long story short, he thinks protein is a better solution.
Three articles in the series all together. Good stuff. Plus I guess we can start making jokes about wooden starships now...I guess neutron activation would be less of a problem...


Alex @ 48, 52: You've just made my point for me. Ask any experienced SWO (Surface Warfare Officer) just how close they'll get to a carrier in bad weather... and if you're at the O-club bar, you'll find out that they don't shy away due to the technology, but due to the command structure and who gets selected to command the damned carriers.

Conversely, they have little difficulty with getting near nuclear attack subs, which are in many ways more complex than aircraft carriers... but don't have the same hidebound command structure, relying instead a great deal more on teamwork and a more team-oriented decision process. Even though there's still a captain of the boat, he (thus far, always he) hasn't been "raised" in a system that worships the captain as a deity, and thus seldom acts that way.

Now expand this problem to situations that involve tradeoffs and allocation of maintenance/troubleshooting resources (on the fly, naturally) among life support, navigation, propulsion, etc., in an environment that is orders of magnitude more hostile, and in which there is no rational possibility of finding a friendly port for repairs...


Alex@49 It looks to me like a black hole drive (or something functionaly equivilent, like a zero point energy drive) or a reactionless drive are the only remotely reasonable (assuming someone doesn't figure out how to make wormholes) alternative to starwisp travel. If mass mostly isn't a problem, then closed ecosystems can be a lot less perfect - you take your 8 cubic km asteroid, stick a 3 cubic km iceteroid onto it & build a large colony inside the asteroid. Between having space for large amounts of diverse life, all the free energy you might need, and a truly vast amount of raw materials, and that fact that a drive like this will let you go moderately relativistic (and you've got lots of mass for shielding), and the result looks (to me at least) like a viable alternative to starwisp microtravel. Best of all, if your closed ecology is actually only 99.5% closed, you've got plenty of supplies for the decade or two your trip will take. Of course, this options looks a whole lot more like a hard SF version of one of Blish's flying cities than it looks like any sort of "ship".


@ 68 during the "golden age" of SF?
Well, I was alive and reading SF at the end of that period - when Clarke was still writing shorts, and JWCampbell was editor of Analog ...(note)
And the Earthlings WERE BUILDING space rockets, Sputnik was launched, the BRITS had a small VERY SUCCESSFUL space programme ......
And everyone, us, the US etc THREW IT AWAY.

Weep for the lost opportunities.

Note: first SF I consciously read was my father's pre-war penguin/pelican copy of "Last & First Men" at about age 9.


heron61 @74: you are assuming that such propulsion technologies are (a) permitted by physics, and (b) can be built.

I'm trying to explore the possible options open to us if (a) it turns out that known physics doesn't contain any loopholes and (b) higher-energy physics isn't accessible to us with any tools significantly smaller than the LHC.

Here's the rub: we haven't directly observed violations of known physics or phenomena like quantum black holes or strange matter. This doesn't prove that they don't exist, but it does place an upper bound on their frequency (a universe full of quantum black holes and/or strangelets would look radically different -- in all likelihood, incompatible with human life). It also puts a lower bound on the energy of interactions required to produce them (i.e. the sort of energetic interactions we see in the universe around us today and in the recent post-inflationary epoch is insufficient).

It follows that, even if they are permitted by physics, we may never be able to generate or observe them due to the energy requirements. In which case, using them to propel interstellar vehicles ain't gonna happen.

We could have bangers and mash, if we had any sausages or potatos. But we have neither. So what are we going to eat instead?


Thinking on the huge amounts of energy (as Charlie pointed out) that are required to get a data stuffed coke can upto anything like an acceptable travelling speed, I wonder if we could use the rather large and powerful powerstation we have to hand. Specifically could something like a Dyson sphere/swarm could be used to focus a useful amount of the sun's energy into a form that could propel (and eventually slow down and stop) the coke can -- I've no idea if such tech would require handwavium.

However, given such a system, I *think* the power requirement should allow for scaling to the level required for some small generation ships (assuming that there are sufficient building materials left over after constructing the power focusing equipment). If there is enough matter, I could imagine a system choosing to send out these life seeds (of whatever form) to as many potential suns as they could find.

I wonder what this would look like from the vicinity of a sun receiving such a seed package? I can image an unusually bright burst of light for a period of time, followed by darkness. Or perhaps the sending sun would be pulsing its light between multiple seeds, and the destination would just see regular flashes.


Specifically could something like a Dyson sphere/swarm could be used to focus a useful amount of the sun's energy into a form that could propel (and eventually slow down and stop) the coke can -- I've no idea if such tech would require handwavium.

That's called a Nicoll-Dyson Laser - literally the only tool of genocide on a planetary scale to be invented on USENET by a well-known science fiction editor!


No, I don't think that Hans Moravec errs on the optimistic side concerning the brain's computational complexity. I think his estimate doesn't even come close. The problem is that he treats the brain as a discrete instructionist system. To him, neurons and synapses, clearly a finite number of each, are the fundamental units, and he thinks in terms of the current computer architecture, based on the concept of instruction.

Many in the AI community would agree with him--perhaps not surprisingly, because they mostly study computers, and of the current architecture (NO, it is not the only one possible, although it may be hard to imagine this right now, when we don't have an obvious alternative).

But if you ask a neuroscientist, someone who actually studies the brain and not computers, you'll get a completely different response. Brain is an analog system, whose fundamental units aren't neurons and synapses, nor even individual ion channels embedded in the neuronal membranes (proteins shaped to selectively let through ions of certain kinds, thus creating electric potential across the membranes), nor the action potential events at the synapses (what is often called as a synapse "firing").

The most fundamental factor in the brain is the background "noise" of tiny random fluctuations of electric potential, too small to make a neuron fire. Most of the changes of potential originating in external input are also subthreshold. But, when applied over the "noise," they can cause the potential climb over the threshold, making neurons fire.

Importantly, this is so random as to be essentially unpredictable, essentially a manifestation of the butterfly effect. An arbitrary small deviation will lead to wide swings down the road. And I mean it when I say "arbitrarily small," because we're talking here of a continuum of values (no, quantum mechanics can't come to the rescue here, but it would be too much of a digression to go into this; read Ilya Prigogine; briefly, the spectrum of quantum observables is only discrete in simple linear systems).

Because of this, the brain can "tell apart" between an essentially infinite number of states. In complexity theory, complexity is defined to reflect the number of states the system can tell apart. Moreover, the background "noise" is pink, its amplitude behaving as 1/f with respect to frequency. More generally, a fractal noise is one whose frequency behaves as 1/f^d, where d is the fractal dimension. If d 1, the noise is predominantly "organized" (d = 2 is brown noise, d = infinity would have been an organized structure). But when d = 1, this is the domain of highest possible complexity--reached in the brain.

But you may say, what does it matter if the states the brain can distinguish between are random? And how does that benefit the brain? You may or may not have heard of Neural Darwinism, a paradigm proposed by Gerald Edelman, a Nobel Prize winner, which is gaining support among neuroscientists and philosophers of mind (Daniel Dennett, for example). The brain is a massively parallel computational system, where neuronal ensembles compete for the best fit to the sensory-motor correction loop. Think of this as a gazillion of teams working on the same problem, adjusting based on the continuously coming input from their actions. The winner is rewarded by strengthening of synapses, which are more likely to fire in the future. Etc.

The brain thus works by the evolutionary principle. But there can be no evolution without random variations. This is what pink noise is for--the most computationally dense noise in the universe. The brain uses the computational power of deterministic chaos (singularities are occurring in the brain all the time! every state is like a ball atop a hill, deciding which direction to roll down--yet to find itself on another hill), which is truly enormous, dwarfing the Hans Moravec's estimate based on counting discrete units like a galaxy dwarfs a nanoparticle.

I predict that we will not develop artificial general intelligence until we do the same, developing a computer architecture based not on instructions (this is a dead end), but on the computational power of deterministic chaos.


charlie@74 It is a bit rich suggesting that micro black holes are outside of possibility for a drive when your main article blithely talks about the uploading and downloading of minds for a starwisp like vessel as if that were just extrapolation of known phenomena rather than just faith.

At this point we neither know whether any such black holes have formed and subsquently evaporated or whether larger ones exist but are relatively rare. It may turn out that they exist even within the solar system and 'just' need to be detected and located and driven to the space dock. I wouldn't bet on it, but this is the sort of technology that might be game changing in terms of the discussion.


I believe the journal software has garbled my "less than" and "greater than" signs (and a few words in between). Let's try again, with HTML codes:

If d 1, the noise is predominantly organized (d = 2 is brown noise; d = infinity is an organized structure without noise).


correction to @80. You don't talk blithely about uploading minds, but you do assume it might be possible as you explore the math of transmission vs storage.


As to starship voyages, the situation is not as dire as it seems. The problem only arises with the methods of propulsion that rely on internal resources. If you draw energy from the environment, then things become different.

Leaving the solar system (roughly) in the plane of ecliptic, going past Pluto, yadda yadda, is a time-honored approach. But, in my opinion, it is less than optimal. Here's what I suggest: Build a magnetic sail spaceship on Mercury, then swing it by the Sun to get an gravitational assist toward the south pole, and follow the Sun's south polar jet. (Yes, the Sun does have polar jets, though they are not as strong as in some stars where they glow.)

The polar jets follow the so-called "open" magnetic field lines (called open because we cannot see the end). But surely the end exists (given there are no monopoles); they just happen to be long. I suspect that the magnetic field lines (and the polar jet) coming out of the Sun's south polar goes all the way to Alpha Centauri A's north pole, but where it continues from its south pole I have no idea.

Why magnetic sails? Simple: you get a continuous acceleration for free, as long as you keep following the polar jet. Given that the characteristic velocity of plasma flow in the universe is about 1,000 km/s, with the typical velocity at the center of filaments (the jets) around 10,000 km/s (1/30 of the speed of light in vacuum), I suggest that such a ship could attain the speed close to 10,000 km/s essentially in no time and thus arrive at Alpha Centauri in only slightly more than a hundred years.


OK, garbled again. I shouldn't have posted after preview, since the "less than" and "great than" were already resolved. Just in case, here's in words:

If d is less than 1, the noise is predominantly chaotic (d = 0 is white noise). If d is greater than 1, the noise is predominantly organized (d = 2 is brown noise, d = infinity is no noise).


And speaking of uploads/downloads of mind, consider this: Digitization of an analog system with continuum of values always implies some loss.



I *like* reading about the exploits of starships and their crews, even if it's is entirely not probable.

Starships are here to stay. Brain transmission is boring.

RAH! ;)


#86: Don't worry, Dude from over there. I think brain transmission is rather improbable. Starships, on the other hand, are very probable.


Christ Charlie, I wish Greg Egan would pop out of his rabbit hole in Perth to come to Worldcon next year. I'd love to listen to you two discuss all this.


Kevin Murray @77, Lasermotive may not quite have the speed yet, but they have possible technology.

Alex @78, James is a reviewer and first-reader, not an editor.


Nestor@70: nothing like a real brain simulation, just point neurons, but an interesting concept. I thought at first it was a continuation of the Blue Brain project, but it's nothing of the sort: http://www.wired.com/dangerroom/2009/11/darpas-simulated-cat-brain-project-a-scam-top-neuroscientist/

Then again, do we really need to simulate every protein and ion channel? And do we really need to model our AIs on brains that have evolved to control 300+ muscles, or is there a more economic option?

Insects are surprisingly smart for their size: http://www.sciencedaily.com/releases/2009/11/091117124009.htm

(BTW, the relevant researchers' own websites/blogs have links to the original papers)


Leonid: sure digitizing analog systems results in loss of accuracy ... but the important question is, how accurate do you need to be? We lose brain cells at a prodigious rate throughout our life, and create new synapses along the way. We mess with the background action potentials and the soup of neurotransmitters our brain cells live in every time we drink a cup of coffee or inject heroin or eat a meal or do just about anything. I'll grant you the neural Darwinism model sounds plausible -- it's not obviously wrong, and it's less obviously wrong than simple connectionism, but I don't think it invalidates the conceptual plausibility of uploading any more than the lack of an artificial atrioventricular bundle invalidates the design of a heart/lung machine.


Brendan @ 63 - actually, you just gave me a good idea. Instead of transmitting thousands or millions of scanned minds, transmit one composite mind made up of the average of the stockholders/donors/important politicians who support the mission (along with the minds of people with the required skills). It'd require less memory and processing power, at the expense of the mind having 10^6 different memories of its third birthday party.


I don't believe that you could "average" personalities or memories, but how many copies of (for instance) the visual cortex would you need on your starwisp? Only one, I suspect.


Charlie, I think the question is not really the loss percentage; it's not quantitative but rather qualitative. But Prigogine's theorem, any discrete system is predictable, whereas a complex nonlinear analog system is not.

Even if you have an NP-hard discrete system, which means that there is no algorithm that allows to compute it faster than just reproducing the system in its entirety, the point is that it *can* be reproduced and, therefore, a more computationally powerful system can predict it in less time, before it ran through its paces. A complex nonlinear analog system cannot be reproduced *in principle* because of the butterfly effect.

You lose not merely an amount of information in transition from brain to a digital system. You lose the potential to tap into a continuum of distinguishable states for computational and information storage purposes. A discrete set simply cannot be compared to a continuum if all of its values are potentially available in a meaningful way, a way that "makes a difference" to the system.

Which is why I said that we will not develop AI until we follow the same evolutionary paradigm in its design. (In general, evolution has been, and always will be, beating any rational design.)

In other words, the Turing's idea that we can simulate the brain's functionality without knowing how the brain works is wrong, because it is in fact based on a wrong (IMHO) assumption that the brain is essentially an instructionist, algorithm-based system. But if the Neural Darwinism paradigm is right, this is not the case and we may have to closely mimic the brain in order to produce a true AI. I call it the neuromorphic principle in my stories.

As a matter of fact, such an AI will not be a digital system.


Charlie@76: But we don't need new physics, and more to the point, unless it's really new physics, the energy requirements will be the same. Here's a couple of ideas:

The Really Big Accelerator. This is just an electromagnetic accelerator several billion kilometers long or longer. Built in space, of course, and like a real Dyson sphere, not contiguous. If it's big enough, you can bend it into a loop so that an object can be accelerated over several passes. Why not build it with a radius of, say, 50 - 100 A.U.? Nice 'known' technology, the power requirements are relatively modest, and you have something that can fling a mass of anywhere from kilograms to hundreds of kilograms outwards at significant fraction of the speed of light.

Acceleration by Really Big Particles. Same deal as above, only the total mass of the launcher can be much smaller. The idea is that rather than accelerate payload at a few g's, you can accelerate a metal slug at several thousands or tens of thousands of g's without significantly degrading it's performance. Now equip with interstellar transport with a current-carrying ring (or several of them); the metallic slug whips through the middle of the ring and imparts momentum via electromagnetic coupling to the ship. This method is good out to several light-years. Better: you can also use this trick to decelerate at your target destination, and even ride it backwards to get home again. It's even possible to reprovision in mid-flight, provided the contents of the metal carrying shell can survive the acceleration.

Big engineering, big energy and big money, of course. But you're going to need those no matter what. Again, modulo some weird physics breakthrough. But that sort of breakthrough would make table-top fusion as a science fair project done by a couple of bright fifteen-year-olds seem mundane and unremarkable.


So, to wit, I suppose it doesn't invalidate the possibility of "uploading" a mind into some other format. But that format will not be truly digital. Not in the sense we understand it. And the computing architecture to sustain that will be *very* different from the current one.


ScentOfViolets@95: I suspect we don't need to build such an accelerator. It already exists in the form of the Sun's polar jets. Granted, I don't have a proof that it goes all the way to another star, although I have some very good reasons to think it does. But this is a subject worth a separate discussion. Perhaps some other time I'll get into this.


At the very least, we could send a probe down the polar jet. We may even get an idea of where it goes without it having to follow it all the way. And we can already do it with the existing technology. Although the onboard computers to manipulate magnetic sails in real time (magnetic field lines, and therefore the polar jet, will not be straight) would have to be quite sophisticated.


A simple example to show why not knowing how the brain works means we really don't know what we can do about uploads and such. A few years ago I attended a workshop on parallel and distributed computing. The guy sitting next to me was a neuro-anatomist who had mapped the entire nervous system of the sea slug. Turns out they all have 97 neurons, and the neurons are all connected the same way in all sea slugs. Nevertheless, the sea slugs have more than 97 different kinds of behavior, and the anatomist had shown that many of them use the same neurons in different ways. So there must be other characteristics than connectivity that determine the behavior of neural circuits. Now scale that number of neurons by a factor of 109 to get a human brain. Will finding out how the neurons are connected tell us the complete state of the central nervous system? Doesn't seem likely.

Eric @ 3
Drexler has done a lot of work on the feasibility of machine-phase assembly in the last few years. Check his blog at metamodern.com for pointers to replies to those objections about quantum uncertainty in placement (hint: you don't have to compute placement, you use a jig, just like any machinist in the macroworld), plus ideas (based on recent experimental work) on how to use self-organization of DNA in machine-phase.

If you've lots of energy available to you there are some other kinds of launchers that might work. The most elegant one is just a torus of degenerate matter massing a few trillion tons. Spin the torus up to very high rotation and fire the payload past the torus so that the advancing edge accelerates the payload through its gravitational field. Since the payload is in free fall the whole time its not affected by the acceleration, and its trajectory can be shaped so that as the leading edge of the torus moves back it's too far away from the payload to decelerate it significantly. You can see why tidal forces are a concern: the gravitational gradient needs to be really huge to make this work.

Of course if you have a handy neutron star you can use a gravitationally-assisted slingshot maneuver to get a similar effect, but I don't see one nearby.


Okay, I'll bite. Here's your "minimal handwavium" solution. (I use only one handwave instead of two, and it's a very small handwave on something that already exists and has a clear path to success.)

We can already accelerate something to solar escape velocity. Thus lobbing a package at another solar system is easy if we don't care how long it takes to get there. No handwave necessary.

We don't need fuel-heavy or energy-heavy solutions. We can use a minimally powered laser/solar sail system to launch lots of little packages at low velocity. Since this solution demands nothing more than basic solar escape velocity, a single nuclear reactor on the moon (or any asteroid) will power as much laser as we need. Since we don't care much about velocity, we'll launch two-litre bottles instead of soda cans. No handwave necessary.

Since we're launching at low velocity, we'll use atmospheric breaking. (And possibly be decelerated very slowly by the interstellar medium) We don't care about final approach navigation because we're launching thousands of capsules, and gravity will eventually attract our capsules. We might use some kind of glider configuration, or we might allow our package to deploy a parachute.

Here's the handwave. Each package is a genetically engineered Von Neuman machine. Since we're already doing limited genetic manipulation, and already have the tech to do more once we fully break the genetic code the handwave is minimal. (If necessary, we'll write our own genetic code instead of trying to decompile what nature gave us.) However, if our knowledge of the subject simply continues to grow at the same rate it's been growing, we should be able to launch our first capsule in 100 years or so.

As soon as I thought this through, I realized that I'd just rewritten "The War Against the Cthorr" from the alien perspective. (I wonder whether Gerrold built his "starship" solution first, or decided to write an alien invasion novel and then invented his landing vehicle?) But since it's been hinted at a couple times in our last few discussions, I decided to post anyway so it would be explicit. I also note that we could do this now if we simply loaded the capsule with some fast-growing algae and a little kudzu. Season as needed with whatever bugs will survive freezing, and launch.

Charlie has previously discussed the ethical issues with various possible solutions to star travel. This particular solution is very, very unethical, of course, given what it will do to anything that lives on another planet. What I find interesting is that the lower the energy needed to get something earthly to another star, the lower it is on the ethical ladder...


Given all the issues raised on this thread and others, the most likely form of extrasolar colonization would seem to be small, self replicating, seed ships.

Send small hibernating AI ships to target systems, so speed doesn't matter. Once they're there, they can spend centuries building the necessary equipment and mechanosphere. Then terraform an acceptable world, or modify your biological templates to survive the biosphere there already. Construct your colonists and raise them with androids when everything is set.

It would be attractive for the ideologically motivated, since the mission designers would have absolute control over the memes transmitted to the colonists.


It would be attractive for the ideologically motivated, since the mission designers would have absolute control over the memes transmitted to the colonists.

It would work until some fool misused his/her spellcheck and spelled "bastards" as "buzzards," resulting in a war against the raptors of the motherworld. Or something.

I'd also wonder how long it would be before some ideologue got us "arrested" by the Pan-Galactic powers for throwing Chtor Bottles at inhabited worlds.


So, in other words: burn the ships, blow the dandelions?


The spore solution is a really common one in nature, a spore being defined as a really tiny (typically one-celled) package launched to colonize some remote place. Things up to the size of tree-ferns use it, and it's one of the preferred ways for organisms to colonize remote oceanic islands. Bacteria have been using it for billions of years, too.

What we're talking about is the civilization-sized version of a spore. That's fine.

The one point I'd make is that spores work best when they're colonizing ground with few or no competitors. If they land in an area rich with competitors, they work best when they can form a symbiosis with the locals, or parasitize them. This is the trick that orchids use, as do lichens.

Scaling up from an Earth biological system is always interesting, but given that we have little or no idea what's in another star system, taking over an Earthlike world from a spore the size of a coke can (or even a large ship) is probably unworkable, even if the planet doesn't have intelligent life. A better target is either an asteroid or a comet. The spore can use the raw materials to build an intermediate that can go on to colonize the rest of the system as feasible.

I suppose it might be feasible to take over a technological planet. "We are of peace. Always." is probably a good entry line....


Facinating work. It's interesting that modernist writers, especially Pound and Stevens, saw their project as a "ship" at sea. I think that the starship cannot contain the same mythology because interstellar travel is so difficult. Perhaps people intuit that the posthuman explorers that can achieve this will be nothing like themselves. So, SF is at a kind of stall because the world of Stross is not like Star Trek in which fleshy humans seek the "other"

In seeking out new life and becoming posthuman in the process, we become the "other"


>every time we drink a cup of coffee or inject heroin or eat a meal

There's a Stross quote of the day. I know you're in Scotland Charlie but surely the junkies prefer to shoot up in the alleys and not the local cafe?



Both zero point energy and (especially) reactionless drives are clearly bat-science. However, if Hawking is correct, then the lack of quantum black holes is simply due to the fact that small ones evaporate somewhere between a few 100,000 and a few billion years after they are created. Also, even if they don't naturally occur, we might well be able to make one. I'm not certain that I regard that possibility as any less difficult or likely than the level of nanotech needed to allow a single starwisp to (admittedly quite slowly) build a functional technical civilization. Neither one breaks any known physical laws, and both are well beyond our current capabilities.


heron61: the trouble with building black holes is that, much like making anti-matter, containment is hard and the initial energy input is astronomically high.

If we could manufacture circa 1 megaton black holes and keep them in a magnetic bottle, then: relativistic interstellar travel becomes feasible.

Ditto if we could manufacture multi-ton lumps of antimatter and keep them in a magnetic bottle: Bussard ramjets don't work with a pure fusion cycle, but if you use them to provide the reaction mass for an anti-matter powered propulsion system they may offer another way to relativistic interstellar travel.

The trouble with either method is ... well, the energy output from annihilating 0.5Kg of anti-matter with matter is roughly equivalent to a 21 megaton nuke. This should give you some idea of its energy density. And the process of synthesizing anti-matter in a particle accelerator is horrifically inefficient: the most that's ever been made at a time is on the order of a few hundred atoms of anti-hydrogen, the creation of which required a particle accelerator sucking megawatts of juice.

Black holes: you really don't want to be around a quantum-scale black hole below the 1Mt mass range -- they evaporate really fast (the last single-digit thousand tons of mass blows off in a few milliseconds, and again: we're talking 21 megatons of explosive yield per Kg). You've got to shove thousands of tons of mass through the eye of a lepton-sized needle in a matter of microseconds if you want to stabilize it -- or store it in a relativistic storage ring. Storage rings are great for holding particles up to a few GeV (travelling close to the speed of light so that relativistic effects slow their decay as viewed from a resting reference frame), but for a black hole we're intent on stabilizing we're not talking GeV; we're talking eV with 20-30 digits after them. That's going to take some serious reinforced concrete foundations (and then some) ...

The statement "It's permitted by the laws of physics" is not isomorphic with the statement "It's easy!"


For the initial launch of the seed to the star's, be it a generation ship or a can of goo, one can't look past the biggest reactor on the block, the Sun. The only problem is the energy output is to diffuse when you get further out into the solar system which is where a close solar orbit solar panel array like in Ian Wallace's Rape of the Sun would come in. It is with in our current capabilities to build and setup, the only problem with it is that it would be a rather effective weapon for cooking off life on Earth with some mis-aiming of it's beam.
Or there is the Crazy option, if we had the will and the technology to convert the entire solar system into one scary as hell spacecraft and go visit the nearest system. If we find suitable materials there we build another ship split our constantly growing population between them and move to the next nearest two stars rinse cycle and repeat.


Here's a little more about the phase diagram for carbon, and the regions of instability for diamond: http://dao.mit.edu/8.231/carbon_phase_diagram.jpg

Of course, the other important piece of information is the velocity of the phase change, which I don't know.


Maybe its just a matter of scale.

Where does a civilization have to be on the Kardeshev scale to allow for fleets of near light speed interstellar manned ships whose cost in energy terms is proportionally equal to the amount of energy currently used by oceanic shipping of a proportional number of people and amount of freight?

IOW, if the current total world energy use is X, what portion of X is currently used to move oceanic shipping? Out of 6 billion people how many are currently at sea at any given time? How many tons of freight are moved each year via oceanic shiping.

Now convert these people, and proportional tonnages of freight (along with the cost and weight of their life support systems, etc.) into starship equivalents. Assume a standard crew complement of 1000 per starship. So if we can estimate the energy costs to propel such a starship to 0.9C, how high up the K scale do we need to be before a level of intersteller shipping equivalent to current oceanic shipping (as a percentage of the civilization's overall energy usage) is achieved?


Does it have to be a laser or maser propelled sail? According to my copy of the "Starflight Handbook" the solar version can achieve about 0.01c while the laser propulsion could achieve up to 0.10c.

That is, if you don't mind going 10 times slower with travel times increased from decades to centuries, but you save a lot of cost and energy. Breaking also becomes easier, just use the photons from the target star.


Maybe we won't have to make giant leaps to actual stars. Maybe we can take small steps to brown dwarfs between the stars. Who knows, ther may be 100s of BDs betwen the stars for every visible star:


"The other headline would be the discovery of a brown dwarf that is even closer to Earth than the nearest star, the Alpha Centauri system at 4.3 light-years. Brown dwarfs are objects that form along with stars but do not have enough mass to trigger or sustain nuclear fusion. They are so cool and dim very little is known about their distribution in the galaxy"



"What if space is littered with these failed stars, scattered between the bright ones like a stellar Polynesia, making interstellar travel a series of short hops, rather than a single gigantic one? What if a simple fusion reactor carried just enough fuel to push a spacecraft to our solar system's Planet X in reasonable time? What if it could refuel there, harvesting just enough hydrogen or deuterium or helium to limp along to another dark neighbor, and another, and another? Granted, it would take a long, long time to get to Alpha Centauri that way, and probably a much, much longer time to find a planet somewhere that looked even remotely like our rain- and sun-drenched Earth. But given the likelihood of tidally warmed moons, and the obvious possibilities for life there, we may just find that the cold, dark spaces are where most of the action is anyway....Discs around brown dwarfs typically weigh about one-tenth of the mass of the star itself, so in this case it probably contains one or two Jupiter masses of available planet-building material. "I'd speculate that it could build a Saturn, or maybe a few smaller Earth-sized planets," says Luhman. What is more, these would-be planets could be habitable. The surface temperature of the mini brown dwarf is about 2000°C, which means that any planet 1.5 to 7 million kilometres away could maintain liquid water. The disc probably straddles this range. Luhman hopes to find out whether even smaller objects - perhaps as little as five times the mass of Jupiter - can reign at the centre of nascent planetary systems. "It's still an open question as to how small you can go, but hopefully we'll be able to answer that soon."

There may be dozens or hundreds of mini-solar systems between Sol and Alpha Centauri. With the discovery of brown dwarfs, free floating planets between the stars, and extrasolar planetoids like Sedna, future space explorers may find plenty to keep them occupied in our own solar neighborhood for centuries to come. While not the galaxy spanning empires and federations of science fiction, it would be enough for our species to explore far into the future without the need for exotic starflight technologies.

Now if Brown Dwarfs turn out to be scattered by the dozens or hundreds in the space between the stars (and if most of them have mini solar systems capable of supporting life because enough heat is generated by the BD to allow liquid water and photosynthesis based on infrared frequencies), then the old galactic space operas become as obsolete as dinosaurs on Venus. Perhaps we'll find that our kind of life, based on visible light spectrum photosynthesis, is the rare oddity and infrared based life far more common.

Since these solar systems are a stone's throw away, they can be reached without exotic warp drives or hyperspace. Simple laser sails or nuclear rockets will do just fine. Exploration missions can visit and return in a matter of years, instead of centuries or millennium. Interstellar "empires" and "federations" can be created using slower than light space travel. Maybe Capt. Kirk and Obi Wan Kenobi wouldn't be impressed, but no author has to violate the laws of physics to tell a good space adventure. An "interstellar" federation can be created just with the BDs near to Earth, providing more than enough grist for good story telling. Only the scale has changed, a BD federation would consider Alpha Centauri to be as far away as Capt. Kirk considered the Andromeda Galaxy.


Denni@90: Mimicing the brain's "design" conceptually doesn't mean simulating every single ion channel.

As to the brain having evolved to control movement, you're absolutely right in saying that. But I invite you to take a further step and say, following Rodolfo Llinas, that our thinking, although many degrees removed, is ultimately an internalization of our movement, because the universals of the external world are embedded into the brain's structure (via evolution) through movement.

A "brain in a jar" (to wit, something that doesn't have an action-feedback-action loop with the outside world) is impossible. Any any interaction with the world means change, which ultimately is a movement of some sort. The more complex, the better. Consider that our brains are more advanced than those of the other animals in a large part due to the far greater number of degrees of freedom in our movement, mostly thanks to our extremely agile hands.


By the way, for those authors willing to violate the speed of light barrier: Consider the modified Lorentz relativity.

The principal difference between Einstein and Lorentz relativity is that in Einstein's, speed affects time and space themselves, whereas in Lorentz relativity, it affects only clocks and rulers (the ways to measure time and space). In Einstein's, there is no preferred frame of reference, and hence no absolute time and space. In Lorentz's, there is a preferred frame of reference, and hence absolute time and space (in the original version, Lorentz invoked the concept of aether as the preferred frame of reference; in the modern, modified version, the preferred frame of reference is the local gravity field). In Einstein's, the speed of light in vacuum is the absolute limit on speed, whereas in Lorentz's, there is NO limit on speed.

No experiment to date has been able to distinguish between Einstein's and Lorentz's relativities. Noteworthy, the GPS system uses the Lorentz relativity, for the sake of simplicity of calculations. See, for example, this:



Instead of Anti-matter, what about Tipler's Baryon Tunneling for a power source? Not that I understand anything about it! 8-)


Looks like this post got a nod on Reason.com:


You don't need no fancy new technology - all you need is to drop the ridiculous "in my lifetime" requirement. The unablity of considering goals beyond one's lifespan is a rather unique and deplorable feature of the contemporary North American (and to a lesser extent general Western) civilisation. Historically, such short term thinking is the exception, not the norm, as testified by countless monuments which have been errected over centuries.

And on a personal level, many people have risked and are willing to risk their lives to bequeath their offspring a little patch of land - here we are talking about a whole new world!

Travel time does play a role of course, but a long as it is short against the expected lifetime of the destination, it is just another mission parameter to optimize. Energy-wise, the sweet spot is roughly at the point, where the total energy for life-support is in the same order of magnitude as the energy used for propulsion. For the mass budget in the case of a reaction drive, this translates into trade-offs between reaction mass (propulsion only), leakage rate of biosphere (life support only) and energy carrier stockpiles (both).

A generation ship with a dry mass of 1 Mio. tons (about the mass of an oil rig) and a cruise speed of 300 km/s is well within the capability of nuclear powered electromagnetic propulsion and can reach a target 20 ly away within 20000 years or 800 generations - no fusion, no nanotech, no 100+ km megastructures, no hard AI, no magic wand. The ship can be manufactured in space over centuries by the very same people (tribe would probably be a better word) who will evenually fly it - this is in fact the only reasonable way to do it: As maintaining the ship over millenia is tantamount to continuously rebuilding it anyway, you might as well have it constructed by the crew - the few hundred years this might take are negligible compared to the transit time and constitute a valuable test and training period. The generation at launchtime will have been born on board and regard the ship as their home.


As for communication: If you have solved the problem for 1000 AU, then you have solved it for practically anywhere. Simply send a relay station into the opposite direction of where you intend to go and then use the sun as a gravitational lens to focus your transmission.


Andrew G: Looks like IEET's somewhat excessive reprint got a nod from reason, not this blog. Cough, cough.


This entire line of argument has very interesting implications for SETI and the fermi paradox. If the first thing any civilization interested in colonizing the universe does is to miniaturize their technosphere (love that word) into a package of nanotechnological tools, and acclimate themselves to life in space by becoming part of said package of tools, then "where is everyone" may suddenly have a.. rather.. absurd. Answer.

IE: "Everywhere". The solar system could currently be home to hundreds, or thousands, of alien polises , and if they did not want us to know about them, we would not be able to detect them. 30, or for that matter, 3 billion uploaded alien space bats living in a cavern excavated in the sea of tranquility 700000 years ago would be easy to hide.
Of course, untestable theories are drek, so what would be the give away signs that this is the kind of universe we live in? Hmmm.. Waste heat from large communication laser arrays near interesting star systems and astronomical phenomena? Of course, being lifebearing, the Sol system would qualify as "of scientific interest", so there really should be such arrays here. or their remains.


I think perhaps we are assuming that the total data storage to transmit a human mind is high. I'd postulate that there's a very direct way of capturing the a reasonable reading of the state of a human brain in use today: the MRI.

Also, the advance of MRI technology is a necessary one in the assumption of any mind uploading/downloading. In order to upload a physical meat brain to data, you need to be able to read the total state of that meat brain, which is effectively MRI technology.

So a better estimate of the total storage necessary would be represented by a more up-scaling of the total data storage requirements of a modern MRI brain scan. If we conceptualize this as some sort of "HiDef MRI" then a conservative estimate would be multiplying the total storage of a modern MRI brain scan by a factor of say 1,000X, with a ballpark being more like 100X, and given futuristic data compression 10X.

Lacking a working knowledge of the total storage that actually takes, we can make assumptions based on the limits of known technology. So if we assume that the MRI machine's hard drive has a 1TB hard drive, max it would be is 1TB, or 10^7. More reasonably, we're looking at something


Thomas: there is the countervailing argument that, with respect to the Fermi paradox, there is no paradox: we've seen the ET civilizations, but they operate on such a huge scale we've mistaken them for gross cosmological phenomena. Robert Bradbury, who came up with the Matrioshka Brain concept, points to the existence of globular clusters in the halo around our galaxy full of reddish, metal-depleted stars, almost as if someone's been mining them ...

For that matter, what do we really know about dark matter? :-/


Jason@123: MRI gives some useful information, but it doesn't represent the mind, not even close. It shows us the structure of the brain but tells us nothing about neuronal activity.

fMRI does tell us a few things about neuronal activity, but indirectly and fairly coarse-grained. It is impossible, for example, to tell whether fMRI picks up action potentials in pyramidal cells or in the neighboring interneurons (the difference is as between excitatory and inhibitory activity). But what's worse, it is the spatiotemporal *patterns* of neuronal activity that are important in the brain, not the rate of firing, and the patterns can change without affecting the rate. Whereas all that fMRI can pick up is the overall level of neuronal activity (where there is more oxygenated hemoglobin in the blood = more energy required by neurons).

For more detail, I recommend Rhythms of the Brain by Gyorgy Buzsaki. He does a very good assessment of every method of measuring neuronal activity that we have available (in the early chapters), analyzing both the advantages and the disadvantages (more of the latter). Even direct targeting of a specific neuron (yes, it is possible, though extremely hard, to nail a specific neuron out of the billions!)--well, that doesn't tell us everything even about *that* particular neurons. There are good reasons why.

Most of what we do with the brains can be compared to judging the state of the liver by skin color. Sure, too much yellow may indicate a problem. But is it all there is to know about the liver?


Not to mention that fMRI doesn't tell us about the potentials in the synapses and across the neuronal membranes (never mind any individual ion channel). It doesn't even tell us such a coarse indicator as the levels of various neurotransmitters in different locations, nor *how* they are acting this moment (which may be in very various ways). It doesn't tell us much about the gliotransmitter activity either. I could go on.


@124: Actually, here's a question for the experts. I don't know the average spacing of G2V stars in our galaxy, but we need some data on the average distance between G2V stars.

Here's the Fermi Paradox question:

Let's assume that every G2V star has Earth around it, and they're all the same age.

The question is: could we detect our exact equal around the nearest G2V star?

If the answer is no, then the explanation for the Fermi Paradox is that our detection capabilities are inadequate to detect the one kind of technological civilization that we know exists.


Heteromeles@127: How would we look from space? A triple system with one star of a moderate size (the Sun), one small brown dwarf active in the radio range (Jupiter), and a very tiny body, also active in the radio range and rotating very close to the star.

This is assuming that our radio emissions aren't swallowed up by the interstellar medium. This is called the plasma frequency effect: plasma is transparent to electromagnetic radiation above a certain cutoff frequency (called the plasma frequency) but absorbs it below the cutoff (and then emits it as a black body, "thermalizing" it).

One has to be a fairly strong radio source to get through this. What, did you really thought the space was vacuum?


Heh, I reread what I wrote and it made me wonder: would the alien astronomers think that the Earth is a black hole? One would have to possess a very strong gravitational field to produce the kind of radio radiation we do by accelerating matter from the neighboring star (the Sun).

Storage rings are great for holding particles up to a few GeV (travelling close to the speed of light so that relativistic effects slow their decay as viewed from a resting reference frame), but for a black hole we're intent on stabilizing we're not talking GeV; we're talking eV with 20-30 digits after them. That's going to take some serious reinforced concrete foundations (and then some) ...

The statement "It's permitted by the laws of physics" is not isomorphic with the statement "It's easy!"

I think the salient blow-back is "cost". Never mind the exact mechanism, just put it into a black box. Now just look at the cost of the power requirements. Suppose you want to boost a probe up to 0.866 c; at this speed the kinetic energy is the same as the rest mass energy, which makes other calculations a bit easier. To boost a kilogram up to this speed is going to take about 9x10^16 Joules, or about 2.5x10^10 kWh. Assuming $0.05/kWh, that's $1.25 billion dollars per kilogram of mass boosted up to speed, assuming perfect efficiency in your black box drive. How big a piece of matter are we talking about for the ship proper? I'm guessing on the order of tens of kilograms, at least - and possibly a lot more. So you can round off and say that just the cost of electricity to run this thing is on the order of a fat trillion dollars.

That's fine, not a big deal. The real question is, what kind of society can afford this kind of electric bill? You either get some incredibly totalitarian social setup that is still rather well off compared to the Western world is circa 2010 . . . or you get an extremely well-off world where spending this kind of moolah is no big deal. It's like the joke about Jesus turning Lazarus into a loaf of bread: when the family and friends shout to the rooftops about the miracle of Lazarus' rising from the dead, Jesus points out that the real miracle was turning the guy into a lump of dough first. The rising part is just what bread does.

So the real trick in writing about believable "starships" is not rhapsodizing over physics pron, it's about writing believable societies that can launch them. You're essentially talking about a world where the cost of energy measured in kilowatt-hours is on the order of hundredths of a cent, or less. A world where gasoline(or it's equivalent) is not $5/gal, but $0.05/gal, or $0.01/gal or less. And that includes the cost of cleanup for burning the foul stuff. This is a world where a family of four living in anything less than a twenty-room pile is considered to be living in abject poverty. And so on and so forth.

As for any sort of embellishment on how the ship is propelled? Hey, we've all read the same books, perused the same magazines, taken the same physics courses. You need about a paragraph of exposition for that part.


@128: No, Leonid, I appreciate your attempt to estimate my ignorance, but it's a bit off.

The problem is that radio strength drops off as a square of the distance. A few years ago, NPR's All Things Considered gave a demonstration of what the Earth would sound like at radio frequencies a few light years away. It was the universal background hiss. An astronomer pointed out (helpfully) that Earth's radio broadcasts would be indetectable to anything except the most sensitive instruments at Alpha Centauri. This is because, as you point out (helpfully), interstellar space is not a pure vacuum, and the frequencies we broadcast at the loudest get swallowed up in the background noise faster than you think.

So, this is essentially a math question:

1) How loud/bright is the Earth at a given frequency?

2) How fast does that frequency attenuate in the solar system, and in local interstellar space?

3) What is the average distance between G2V stars?

4) What is our ability to detect said frequency, at Areceibo or any other appropriate facility?

Given the NPR example, I suspect the answer is that we couldn't detect our type of civilization around the nearest G2V star (which, so far as I can tell, is Beta Canum Venaticorum, 27.4 light years away).

Note that I'm not ignoring the possibility of some other type of radio-visible civilization. I'm just saying that we might not even be able to detect ourselves from a reasonable distance.

The Fermi Paradox may be simply due of the limits of detectability and the difficulty of interstellar travel.


Actually, other than being a binary star, the Alpha Centauri system is pretty damn like ours. A is something like 1.1 size of the Sun. B rotates around A at about the same distance as Uranus is from the Sun. So if there is a planet in an Earth-like orbit, it wouldn't notice much difference.

I bet even the solar activity in Alpha Centauri is synchronized with the Sun's, with an obvious delay. I wonder if anyone has done such research.


Charlie @ 76:

Isn't this a little pessimistic? Suppose for a moment that nothing less than the LHC will give us any access to higher energy physics... is this really the obstacle that you make it out to be?

Since it's late here and my brain isn't working so well, take "antimatter" that we always hear about in bad science fiction (usually TV, sometimes movies, less commonly in print). People like to crow about how at best it's just a highly dense energy storage, that the "fuel" takes much more energy to create than actually exists in the resulting product. This seems true at first... but has anyone ever ruled out some sort of exploitable breeder reaction?

With fissionable materials, there are such scenarios. You not only create energy, but you can create more fuel than you started out with. I've never held the opinion that this is a fluke, that it doesn't apply to starship Enterprise fuel as well.

And as for the really interesting stuff, the higher energy physics, is this scenario any less likely? What if it's only difficult to initially tackle, but once you manage that you can bootstrap yourself to much more efficient (and smaller!) devices?

You claim that this seems unlikely because we don't see evidence of it with our telescopes. But I just can't buy that... it's like telling me that fission breeder reactors are probably impossible because I can't point binoculars at the sky and see evidence of it in Boeing jetliners which would just dump the excess uranium from trapdoors in the nuclear turbines. Would there really be all that much evidence of it in the nightsky, even if we aren't the first species to make it that far? (And I can't rule that out either, the universe is what, maybe 14 billion years old, and the first part of that didn't even have many heavy metal solar systems?)

There's a trick out there somewhere that will allow us to do crazy things. I have this nagging feeling deep in my gut that that is the truth. It's one of only two such, the other being that I was born just a few centuries early to get to see it for myself.


The problem with sending out the million-year probes, whether generation ships or technocans, have been touched on by Alan Coren, writing about the Pioneer 10 plaque: I've put up the article here. Briefly, though it's worth reading it all, the catch is that we're doing this because we want our civilisation to be immortal - and that's our civilisation, with people looking like and thinking like us, like the Trek. Give it only 100,000 years and either or both of the people sent out or the people left behind will have evolved into something we have no real similarity with and very little interest in propagating. As Coren says, "So, then, what might be the outcome of all this? The things from Planet X, beside themselves (unless, of course, they are already built that way) at the thought of intelligent life ten billion billion miles away, will hop/crawl/bounce/ drip into their own spacecraft, and set out on the long trip Earthwards. A thousand million years later, and sick of tinned food, they will chuck open the hatch and hurl themselves onto the surface of this planet, gabbling: 'Take us to your leader! Where are the women with the big jugs?' And we Earthlings could, by then, be itinerant warts, we could be chattering fungi, we could all be members of the Labour Party National Executive. What will happen to this already battered old globe when the aliens discover the truth, I dare not even begin to imagine."


A concept is missing in the article or posts (unless my late night reading missed some)

Pier Anthony's Cluser series concept of achieving near instantaneous space travel (it would be both) via mind exchanges with host alien intelligent species in the galaxy-in this case only a willing host can allow such a visitor within its mind (though Anthony explored the possibility of of unwilling hosts and 'possession.' A similar concept is shown in the Stargate TV series storyline with special 'stones' that allow a two way with between human being. (In both cases the concept of sexuality means a person can only transfer to same sex because of some body/mind image clause.) In other words rather than having to recreate a mind and body with this computersese 'upload/download nonsense, just use (diplomatically) the existing mind/bodies out there.

BTW since the topic of the article is the misuse of words or how word choice influence our thinking, I see the same trend in the considerable amount of computer jargon and concepts presented in the article and many posters here. That is very limiting: why do many here presume that the 'mind' should be uploaded like binary bits of memory states of on/off bits? The mind is always changing and SO dependent on our senses that it would like by sending a snapshot of the electrical activity inside a TV in order to broadcast a TV show.

The 'ETHER'NET: I think in plausible terms for the nearer future something simple like a waldo real time connection to space/planetary machine probes could play host to allow a human operator jacked in to 'experience' the sensory input and actually do some real work in space (like build structures, run mining equipment and/or terraform) without ever leaving earth. It becomes a simple -well defined- technical challenge of achieving 'real time' communication in space with quantum internet, ST 'subspace' channels, or 'ether'net as I call it.


Another thing came to mind: we are positing on the absurdity of the words starship but what we are discussing is the limiting semantics of the concept of 'space travel'. Both words limit us to simple 3dimensional space and a very primitive concept of 'traveling' as in a duration of movement in that space.

Traveling as in Pilgrimages , Grand tours, Cruises, Tourism are essentially valid though medieval, victorian and 20th century . Those are the concepts of human experience we must free ourselves from in order to leap to a higher sphere of understanding of FTL. As Buckaroo Banzai said 'no matter where you go, there you are.' We have to free ourselves from small space-time concepts in our use of the terms if we want to imagine ourselves in a future where being somewhere else and experiencing other places can happen without 'traveling' . The singularity that some worship I think will be nothing more than the whole gamut of human experience will be swallowed up by virtual realities and the hybrid versions as well (eg. virtually connecting to space probes). So the discussion we have now about space travel (through uploads or starwisp whatever) will appear quaint and so Victorian as said by others here.

Just think how cartoon characters would seem to us if they discussed the fact that a third dimension must exist even if they are drawn on a two dimensional surface.


Skightly off-topic, yet on-topic.....
"Intelligence" / "life" / autonomous systems (a.k.a. "Robotics" )

I wonder.

Try looking at building a system UP from very simple structures (like evolution does) leads IN THIS DIRECTION perhaps?
Well worth a look.


SoV @130: that energy calculation is ... interesting. And provocative.

Go back three centuries and our energy budget per capita was to a close approximation what we could generate ourselves from food. (Yes, windmills. Yes, sailing ships. Yes, horse-drawn transport. Nevertheless, even stationary steam power was a huge leap forward.)

Project forward and assume breakthroughs -- the one most likely to give us the kind of electrical surplus you're talking about would be, in my books, orbital solar power satellites manufactured from cheap thin-film graphene semiconductors -- and yes, lots of stuff becomes possible. But high energy civilizations present their own challenges, and threats. Notably, it becomes really easy to Blow Shit Up. Your hypothetical 1Kg artefact travelling at 0.866 c is isomorphic with a 10.5 megaton bullet. We already have an inkling of this problem with airline security theatre (250 tons of airliner plus fuel and passengers travelling at the speed of a slow pistol bullet is potentially Very Bad News for ambitious architects); a civilization that can afford relativistic travel is going to have to address some interesting social problems.

dalani, I think you're confusing television scriptwriters' effusions with something other than fantasy (and you've picked up a bad dose of mind/body dualism somewhere -- Piers Anthony?). I'm not going to mock you publicly, but I'm not going to take you seriously either.


In my darker moods, I consider that an quite adequate answer to the fermi paradox - Technological species advance untill they reach so extreme a degree of mastery over the physical universe that a single psycotic with a bad hair day has access to tools that can wreck planets, at which point they (and everything near them!) promptly goes extinct.

Heck, even granting that some species might invent adequate mechanisms of social control to avoid this fate, a society which has no extremists or nutters whatsoever also has no particular pressures for expansionism. Science outposts across half the universe? sure. Colonialism? Ehh.. what would be the point?


Of course, that could be turned on its head, and produce a senario where Really Advanced civilizations promtly atomize into groups of a few hundred to a thousand all of which carefully stay lightmonths away from each other so that the inevitable explosions when someone shorts out the zero point generator in a fit a pique doesnt kill too many people..


Of course, that could be turned on its head, and produce a senario where Really Advanced civilizations promtly atomize into groups of a few hundred to a thousand all of which carefully stay lightmonths away from each other so that the inevitable explosions when someone shorts out the zero point generator in a fit a pique doesnt kill too many people..


@ 139
"Wind from a Burning Woman" ??


Sending people to the colonies is another a bit of baggage you get with "ships".

How much information does it require it encode a human? Well, the human genome is about 6x10^9 bits. Lets use this as an estimate of the information required to encode a tabula rasa human or human-simulation. Then you just need to feed it educational material. I'm not sure how to estimate that, but I expect it will require far far less than 10^18 bits.

Maybe you even want to limit the amount of information you send on purpose, to allow as novel a culture to arise as possible. Creation of new culture has been a big payoff from previous colonizations.

Leonid Korogodski, the distributions you mention are easy to simulate digitally. I have used simulation of long-tailed distributions on a number of occasions to good effect, on a digital computer. Your proposed model of the brain sounds similar to Markov Chain Monte Carlo approaches to sampling of posterior distributions (if suitably parallelized).


Mind/body dualism? I was saying mind/body are one. Piers Anthony's premise is mind/body are distinct. I don't place much stock on that premise for the reasons I gave already.

In many cases what passes for Sci-Fi on TV AND written form are fantasy masquerading as SCi-Fi. I see nothing wrong in that since eg. the miracles of supersonic flight would be total fantasy to someone in the lets say mediaeval age and should Milton have wasted his words explaining Bernouli's equation when no such thing existed?


130 - another advantage to Orion: building lots and lots of very large H-bombs is probably the cheapest way to produce stored energy yet possible. A 100-MT bomb is 10exp17 J, enough to accelerate a kilo of probe to 0.866c, and would cost a lot less than $1.25 billion, certainly if you built in bulk. LiD's cheap*. But, as Charlie says, there's no reason at all to suppose that present limits on energy generation are permanent.


Paul@143: Human beings are not a singular organism, we are symbiotes, depending on cooperation from a correct mix of both internal and external organisms for survival. Quoth Wikipedia: Somewhere between 300 and 1000 different species live in the gut. And that's only the gut, without the rest of the body, and most certainly with no mention of ham, salad or rice and everything required for those to grow. Considering the lack of success I've had with not killing herbs trying to take root in carefully arranged pots on my sunny balcony, I wouldn't bet too much on a fully-formed human and the required ecosystem to support it springing forth from an array of bits somewhere in the butt end of space.


Paul@143: Yes, one can simulate *a* power law distribution on a digital computer. But how would it correspond to the real thing?

The problem is, again, in discrete vs. continuum. As I already said, a discrete system is always predictable, whereas a complex analog system is not.

Consider, for example, the gravity--much simpler than the brain, right? Not long ago, a student of Prigogine simulated on a computer a system consisting of a star, a large planet (a la Jupiter), and a comet. The question was: after how many orbits the comet would exit the system because of interaction with "Jupiter?"

If the numbers were rounded to one part in a million (as initial input and during calculations), the answer was 757 orbits. If to one part in ten million, 38. If to one part in hundred million, 235. If up to one part in 10^16, 17. Even though the system was deterministic, it was not predictable *in principle* because of the butterfly effect, unless you have an absolute infinite knowledge. This is what happens in a system of more than two bodies. Never mind the brain.

(And this is also why calculations of the probability of a small rock on a chaotic orbit hitting the Earth are not worth the computer time spent calculating them.)


I think any conversation about "interstellar travel with realistic physics" needs to exclude pixie-dust stuff like mind uploading from the start. If you want to have people on board, they're going to have to be made of meat. If you are happy with just computers, then say so. But we can't even upload a flatworm and have any hope that it will behave even in a generically flatworm way, far less in a way recognisable as specific to that particular flatworm.

As for the risks Charlie outlines - "But high energy civilizations present their own challenges, and threats. Notably, it becomes really easy to Blow Shit Up. Your hypothetical 1Kg artefact travelling at 0.866 c is isomorphic with a 10.5 megaton bullet." - it's not high-energy, it's high-power that's the problem. Having the ability to accelerate stuff to 0.8c isn't a problem if it takes you thousands of AU of runup and a couple of years to do so. To put it another way: I use enough energy every year in my flat to level a large part of the City of London. However, this does not make me a terrorist threat.

Go back three centuries and our energy budget per capita was to a close approximation what we could generate ourselves from food. (Yes, windmills. Yes, sailing ships. Yes, horse-drawn transport. Nevertheless, even stationary steam power was a huge leap forward.)

Alternatively, you're like a writer who in 1700 is writing about powered flight. Nobody would fault you for writing about steam-powered planes. Otoh, would you have anticipated the collapse of agrarian societies and European-style aristocracy, the rise of the middle class, and the demand for education as a right?

Project forward and assume breakthroughs -- the one most likely to give us the kind of electrical surplus you're talking about would be, in my books, orbital solar power satellites manufactured from cheap thin-film graphene semiconductors -- and yes, lots of stuff becomes possible. But high energy civilizations present their own challenges, and threats. Notably, it becomes really easy to Blow Shit Up.

Vinge had a riff on this one where even a simple JK flip flop required several thousand transistors to deal with the possibility of sabotage. You might run a similar riff on how so much of the mass and function of your hypothetical starship was given over to security concerns, with some character musing how much simpler life was in the 20th when it wasn't the case the 90% of anything was there to prevent the other 10% from being misused. In any event, having believable starships would seem to imply believable high energy societies.



another advantage to Orion: building lots and lots of very large H-bombs is probably the cheapest way to produce stored energy yet possible. A 100-MT bomb is 10exp17 J, enough to accelerate a kilo of probe to 0.866c, and would cost a lot less than $1.25 billion, certainly if you built in bulk. LiD's cheap*. But, as Charlie says, there's no reason at all to suppose that present limits on energy generation are permanent.

ajay, 10^17 Joules is the amount of energy you get from the complete conversion of one kilogram of matter to energy. Is this really the case for your hypothetical bomb? Also, what you have here is 10^17 Joules of energy, not kilowatt-hours of electricity. The two are quite different. I'm using a basic utility rate, true, but I'm reasonably sure I'm right to within one order of magnitude of cost.

Note that I'm not saying that cheap energy is impossible or a show-stopper; I'm merely saying that absent some completely unforeseen physics breakthrough, you aren't going to get starships going anywhere very quick unless you have a civilization that doesn't blink at a trillion-dollar utility bill. With all that implies.


SoV 150: First, 10^17J is, roughly, the equivalent of 100MT. It is also, roughly, the energy from converting 1 kg of matter. (The Hiroshima bomb, 20 kT, was the rough equivalent of converting one gram of matter completely into energy.)
See here: http://www.ocean.washington.edu/courses/envir215/energynumbers.pdf

I'm using your figures for acceleration, quote: "To boost a kilogram up to this speed is going to take about 9x10^16 Joules, or about 2.5x10^10 kWh. Assuming $0.05/kWh, that's $1.25 billion dollars"

And, thanks, I do know the difference between a joule and a kWh, and I'm fairly sure I've got the right one.

Your general point's good - but it does rely on some fairly aggressive assumptions about accelerations, speeds and times. If you're willing to accept a maximum speed of just 0.1c or even lower, the numbers look a lot more friendly. Not to mention looking at some figures for projected economic growth.


ajay@151: I'm not questioning the conversion figure per se, I'm just asking if there are actually bombs that convert one kilogram of mass into energy. A quick search seems to suggest that no one admits to having anything of that size. Is this the case that they actually exist in some arsenal somewhere? Secondly, I'm distinguishing between raw energy and usable energy. If you say that there are 100-megaton bombs lying around and that they cost less than $1.25 billion, I won't dispute it. But how much would it cost to convert that 10^17 J (or higher) into usable output? You can't just wave a magic wand and make it so for zero cost; you have to go into details about the equipment, how much it costs, how it's amortized, etc.

That's why I priced the energy requirements in kWh: so I could do a rough and ready cost conversion. If you think that $0.05/kWh is out of line, too much so to be of any use in an analysis, by all means, explain why.


Do we even need a vessel as large as a 1kg starwisp?

First off, we will probably have a pretty good inventory of data on all planets out to maybe 1,000 light years by the time we think of launching star vessels. This data will include full spectroscopic analysis and even fuzzy images of planets, perhaps as good as Mars through C19th terrestrial telescopes. So we are going to know a lot about he "current" conditions of perhaps 40 million star systems. Even for a science program spanning 100 years, we can reach stars within 50 light years and get data back - which is perhaps 10,000 star systems.

Just how small could we make those vessels? - probably many orders of magnitude smaller than a 1kg reference starwisp.

Secondly, with all the nano tech we will have, we might consider that we could emulate nature and make autonomous devices as small as nature does now, say bacterial level.
They can be self replicating and communicate with their neighbors in swarms. So perhaps we only need to launch a gram of nanotech to the stars and let them use local resources to build transmitters or propulsion systems to report back. Even that gram of mass need not be launched as a solid mass, but just each bacterial sized particle launched in a stream which aggregates as a coupled swarm for the journey.

Let's say Charlie is correct in suggesting that we will never build AGI's and we need human minds to travel to do the exploring. Having uploaded minds (or even functioning working copies) living in a pleasant simulation would be a good way to go. Their clock speeds could be set at any desired level, perhaps speeding up orders of magnitudes to have more thinking time as they fly past each star and do their research. At each star they decelerate some portion of their mass to deliver the tech and information to send back their results to earth, possibly even building permanent monitoring systems.

Why would any minds living this way ever want return to the meat world? They could temporarily inhabit [humanoid] robots and wander about on any world, not just those suitable for humans. This is just a mechanistic panspermia that might precede a biological one if desired.

If each star system gets it's own launcher (linear accelerator/rail gun/laser whatever) then star travel for such beings becomes easy. They launch at near light speed and are stooped by the same systems. The minds can run at any desired speed or even stop, so perceived travel time is as long as desired. It will be a post human/robot universe, not a meat one, but so what? The universe could fill with mind and ideas and be even richer than Stapledon's "Starmaker".


There you go again, Charlie. How dare you take away our toys?

Remember: it's called 'Hard' Scientifiction, not 'Accurate' or 'Reasonable'. It's about the butch feeling we get when pondering thrusting our metal avatars deep into Mother Night, after all the girls have been safely spaced with nary an emotion visible from us beside mild contempt for someone who can't accept her death even though its necessity is made manifest by orbital mechanics.

I've read about our explore at least the near stars an hundred thousand times or more, so I KNOW it can be done. It must be done.

What you write here can't be true, because...well it CAN'T.

(Not surprising that in doing my best at imitating the inner life of those most likely to criticise others for 'whining', I seem to be on the verge of shrieking for the waambulance....)


Re: Orion. My understanding (mostly thanks to links in the previous starship discussions) is that as a nuclear drive, Orion isn't terribly efficient. It's capable of producing high accelerations, like a chemical rocket. But in terms of converting mass to kinetic energy, it's well down the slope compared to (just fer instance) a highly-evolved ion drive.



Give me a billion or two and I'll give you a ground-to-orbit million ton starship that'll go all the way to pluto and drop off colonies at the moon and mars along the way while spamming the solarsystem with probes and manned submissions.

We could have built this thing in the 1960s instead of the apollo program.

I'd build this thing from almost entirely from ship steel using existing infrastructure that currenty goes to cargo vessels and cruise ships, so 'billions' of dollars may not be necessary.

Starships are no myth, just require lateral thinking. When you can lift millions of tons into earth orbit approach you can skip alot of the cost and difficulty of space habitats and long duration travel by a pure brute force approach. Yes yes, radioactive fallout would be horrific from conventional nukes, but with high-yield efficiency it gets more manageable. Nuclear isomers are a possiblity. Perhaps leading to clean ways to initate thermonuclear fusion (Even ballotechnics is a possiblity). Mind you is if there was a asteroid on the way to earth, this vehicle would be a our best/only chance, and we'd not worry so much and love the bomb.


We're still looking for a good reason to go to another planet here. Exploration often has a financial incentive, which is not apparent here. The only reasonable solution is exodus/diaspora. Depending on the details of the uploading process, if upload is sufficiently destructive, these people may have to literally die to leave the solar system.

So, we're looking at systems where people may need to die and use all their funds in order to leave. What kind of transportation can they buy?


Leonid@147, conversely an analogue simulation of a digital simulation of gravity will perform poorly. Randomness is a useful tool, but it's not magic. I think we're missing some important trick with AI, I just don't think this is it.

My own take is that the trick is to do with efficient approximation of the Bellman equation (from control theory). The Bellman equation deals with making decisions to maximize a utility function in the presence of uncertainty, and naturally produces a control system that has curiosity (see "dual control"). Well, it would if it weren't so !@#%^ hard to evaluate in all but the the simplest of toy examples.


Paul Glister, who runs the Centauri Dreams website, http://www.centauri-dreams.org/
which looks at peer-reviewed research with interstellar possibilities has weighed in on Charlie's post.



If we allow self-reproducing nanotech as a payload that can reproduce a complete technical infrastructure in the destination solar system, then we automatically have a way to build large energy collection/transmission systems at very low cost. Just drop a seed onto Mercury and wait some number of years while the nanos convert large areas of the surface to solar collectors, storage units, and microwave (or laser) emitters. The total cost is the original seed; Mercury's mass and the Sun's energy provide everything else you need.

If you want more surface area for collection, have the nanos build massdrivers to launch mass into low solar orbit and build collector/storage/emitter arrays there as big as you like up to the limit of their ability to withstand tidal forces.

A similar technique might be usable to build anti-matter factories, but there are a lot more unknowns there, both in creation and storage of the product.


Paul@157: Yes, randomness is nothing special in itself. But when manifesting as deterministic chaos in complex nonlinear systems, it means that there are more states that the system can distinguish between. In fact, there is a continuum of states such that, however close two states may be to each other, they will inevitably diverge over time. This mean a tremendous computational complexity.

What's more, deterministic chaos in the brain likely provides the necessary randomness for the evolutionary process (see above regarding Neural Darwinism).

What's even more, the world around us, though it may be deterministic, is certainly not predictable. Consciousness differs from FAPs (fixed action patterns, which are more complex relatives of reflexes) in that it is not stereotypical. When you're walking in a park, talking with a friend, you don't pay attention to where you plant your feet; that happens automatically (is a FAP)--until you stumble over a root. That's when consciousness is focused to the matter of where you place your feet.

I maintain that consciousness evolved to allow animals survive in an unpredictable world. Fixed algorithms can only go so far (and robots can't survive over millions of years in arbitrarily changing environments as a species). Moreover, consciousness is not an all-or-nothing thing (as most of the neuroscientists are realizing). It is present, to a varying degree, in all sufficiently complex animals, if only to make them functional. Even in humans, it is present to a varying degree in different states.


Gee. So many typos and mistakes. I must be tired.



The suns gravity lens focuses at 550 AU. You can set up a transmitters at the focal points for each star and transmit clearly back and forth with surprisingly low amounts of power.

Now this is interesting stuff because it demonstrates that when we do achieve the ability to create a human presence around a nearby star, we will have ways to establish regular, reliable communications. A second FOCAL mission, one established at the gravitational lens of the target star, benefits us even more. We could, for instance, create a Sun-Alpha Centauri bridge. The bit error rate becomes less and less of a factor:

…the surprise is that… for the Sun-Alpha Cen direct radio bridge exploiting both the two gravitational lenses, this minimum transmitted power is incredibly… small! Actually it just equals less than 10-4 watts, i.e. one tenth of a milliwatt is enough to have perfect communication between the Sun and Alpha Cen through two 12-meter FOCAL spacecraft antennas.

This seems remarkable, but gravitational lenses make remarkable things possible. Recall that it was only months ago that the first tentative discovery of an extrasolar planet in the Andromeda galaxy (M31) was made, using gravitational lensing to make the observation.


The showstopper that I see here is ionizing radiation.

There's a good bit of hydrogen and helium in space. If you impact a gas atom at 0.1c, it's going to act like hard radiation (Relativity says that a relativistic device hitting a stationary particle is the same as an energetic particle hitting a stationary device). Charged particles can be deflected with magnetic fields in principle (the field would have to be extremely strong, and accelerating a magnetic field radiates energy), but neutrals can't.* That doesn't even start to account for heavier obstacles (micrometeorites, grains of sand, etc).

If any form of highly structured matter was exposed to that environment for ~50 years, I would expect to get unstructured matter out. Tons of shielding are probably necessary.

* In principle, a laser could ionize much of the neutrals. A lot of power would be necessary: the wisp is moving at Mach 90,000 and there just isn't time for subtle nudging. Powering the laser and dissipating its waste heat seem problematic.


Perhaps instead of just moving a downloaded passenger from star to star with launch lasers, we redefine "passenger" and "travel" a bit.
Use the laser to boost the bootstrap machinery to make more launch lasers to other stars, along with instructions on how to emit signals in response from incoming data from other lasers.
Map an approximation of your passenger's neurons onto the connections between stars. He'd think slowly, but for a very long time.

100 billion neurons in the brain, 200-400 billion stars in the Milky War...presumably you could double up a bit in the interstellar traffic.


@163: At the conference below, I'd pointed out that the sun's gravity lens focus at 550 AU is a good reason to have robotic spaceraft ready to head out to 550 AU in the opposite direction of any VERY interesting stellar signals, such as those found by SETI.

@164: I think that it was Prof. Geoff Landis who showed that one can protect an interstellar probe at 0.1 c velocity with a renewable plasma shield in front.

I don't have my notes here, but this was when he presented one or more papers and I presented two or more papers at the (2004) First International Conference on Practical Robotic Interstellar Spacecraft. It was hosted by New York University, and some of us also made presentations at the United Nations.

Does anyone have the Proceedings of that handy?

See also

Many of the questions being kicked around on that thread were answered at that or other academic conferences. Many of we presenters do science fiction AND spaceship design at a rather sophisticated level.

I resist promoting my own pet projects here, either on interstellar spacecraft, or on the nature of the human brain.

Bottom line, I agree that Physics as we know it, and Engineering as we hand-wave, is NOT going to move 2,000,000 HUmans 1.0 to an extrasolar planet. My own publications, however, give me hope that interstellar probes much heavier than 1 kg can reach Alpha Centauri in 50 years (fission power) or 9 years (fusion power), the key being that they are relativistic multi-stage rockets. The old textbook formulae do NOT correctly combine SR and staging.

At the UN, having its 50th anniversary near, was where I suggested that UN Centenary be when we target the video returned from probes to interstellar destinations, and laid out the numbers reversed engineered from that for several technologies and nearby stars.

Mr. Stross is right to clean away the deadwood accreted from softer Hard SF dreams, let alone Fantasy pretending to be Science Fiction. There is still a chance, howeever slim, that true spaceships are possible for the near extraterrestrial regime. The hardest part, IMHO, is the Economic models, not the physics. And the biology/sociology for a small crew for multidecadal missions. Not as gnarly as worldships on multi-century missions, but packed full of unknown unknowns.


A proton moving at 0.1c has about 4 MeV of kinetic energy. (gamma = 1/sqrt(1-0.1^2) ~ 1.005, rest mass a bit over 0.9 GeV). Eyeballing the chart at http://en.wikipedia.org/wiki/Stopping_power_(particle_radiation) it looks like 1 mm of aluminium should be more than enough to stop any protons moving at 0.1c. It's cosmic rays you have to worry about, not interstellar gas. Their energy won't be appreciably affected by relative motion.


Paul@158: "missing some trick"

I assume with AI you mean AGI in this context.

So let me lay out a building instruction for general intelligence:
You need a food source, a permanently running virtualized self-model that contains at least the two states happy (connected to food-source and reducing entropy in the system) and unhappy (entropy in the system is rising), permanent real-time monitoring of the actual system state (situated in the Insula in humans) and a feedback loop between the virtualized model and the monitoring system, where the model is predicting a system state and a perfect match with observed system status leads to a self-reinforcing signal called Qualia.

That is just the basic level, enough to master the mirror test, i.e. on par with common magpies and 18 months-old human babies.

Next step you add models for other selves (including the necessary monitoring mechanisms relaying feedback of course), models for general forms of interaction, models for building models, etc.

Note that this is a discrete sequence, powered by discerning the two states "entropy rising" and "entropy falling" and reaching a successive but logical sequence of phase-changes from there.

It happened in magpies, it happened in chimps (including the "model of other" state) and in humans. It can easily happen again.

You see that there is indeed dual control and curiosity built in the system, but it's not meaningless curiosity. Instead it's focused self-interested curiosity, where the focus and definition of self-interest is defined by the complexity level the system has reached.


Everything evolved to survive. There is no other reason. But consciousness is not a substance or quality, it's a spectrum of system properties. Simple aware consciousness is different from self-aware consciousness, being different from consciousness-aware consciousness.

In general:
Ultimately the only entity able of solving unspecified general problems (thereby having a major survival advantage) is one that possesses continously growing (and therefore theoretically unlimited) complexity. This necessarily implies a continously active mechanism to dump entropy (grow, learn, etc.). Any such mechanism essentially is entropy-guided - no other way out. Such guidance could be passive first (passive selection of survivors), it would someday (phase-)change to active entropy-guided behaviour (hunger, pain) and will really take off from there. Human evolution is just one illuminating example of an underlying principle here.


152: there aren't any 100MT bombs at the moment, but there were in the 1960s. Look up "Tsar Bomba".

Good point about efficiency, but I have no idea what the efficiency of an Orion might be - those codes are still heavily classified.


ajay@169: Nope, this particular bomb was supposedly designed to get up to 100 megatons, but in actuality never did:

Developed by the Soviet Union, the bomb was originally designed to have a yield of about 100 megatons of TNT (420 PJ); however, the bomb yield was reduced by half in order to limit the amount of nuclear fallout that would result. Only one bomb of this type was built and tested on October 30, 1961, in the Novaya Zemlya archipelago.[1]

So it seems a bit, er premature to talk about their costs, or the cost of extracting usable energy from the yield. Do you have anything that shows $0.05/kWh is way out of line for estimating energy costs?


J Boss @168: Everything evolved to survive. Minor nit: everything evolved to reproduce. After reproduction, the evolutionary selection pressure for survival is very tenuous indeed -- we're into selection on the basis of kin survival. There's strong selection pressure for humans to survive long enough to see their children reach reproductive age, for example, but beyond that ... not so much.

SoV @170: Tsar Bomba was designed to go to 100Mt; for the actual test they replaced a chunk of the U238 booster with lead because (a) they didn't want to melt the Tu-95 from which the bomb was dropped, and (b) they didn't want the fireball to touch the ground and spray highly radioactive fallout all over the USSR. The projected yield without the booster layer -- 50Mt -- was indeed a close match for the design, so there's zero reason to suppose that Tsar Bomba wasn't a viable 100Mt design.

As for the cost of energy from nuclear weapons: we know the US nuclear weapons program cost on the order of $2Tn to date. (Figure from a back issue of SciAm; can't be bothered to go digging for more accurate data online right now -- got to go out.) From this, and the number of weapons produced (about 30,000), and the average yield (around 50Kt?), and the stockpile of Pu239 left over (200 tons, or thereabouts?), we can guestimate the energy output of $2Tn in joules (i.e. watts/second).


@ 149: "Alternatively, you're like a writer who in 1700 is writing about powered flight. Nobody would fault you for writing about steam-powered planes."

Hmmm ....

"Soon shall thy arm, Unconquer'd STEAM! afar
Drag the slow barge, or drive the rapid car;
Or on wide-waving wings expanded bear
The flying-chariot through the fields of air.

Fair crews triumphant, leaning from above,
Shall wave their fluttering kerchiefs as they move
Or warrior bands alarm the gaping crowd
And armies shrink beneath the shadowy cloud."

( Erasmus Darwin, 1791 )


JBoss@168: I'm glad you mentioned entropy. What you're describing is known in complexity theory as a "complex adaptive system" (the term coined by Murray Gell-Mann in Santa Fe Institute). By Prigogine theorem, these CANNOT be discrete.

I'm not saying that digital computing systems don't have their advantages. Indeed, our computers have vastly improved many things in our lives. But these advantages have primarily to do with the possibility of exact replication and easy troubleshooting. We tend to (want to) design all chaos out.

So the digital systems can be very reliable; you can make a robot repeat the same action exactly in the same way every time. In contrast, the brain never repeats itself. But this (or rather, the reason for this: the evolutionary process in the brain) is exactly the brain's main advantage; this is what makes it a complex adaptive system capable of dealing in context with arbitrary situations.

We must develop a completely different computing architecture, one that incorporates deterministic chaos (and thus is capable of supporting complex adaptive systems) if we ever hope to achieve artificial (general) intelligence.

P.S. Only general intelligence is worth discussing. The rest ain't worth the name.


Jonathan@166: I still maintain (see some of my posts above) that the best way to reach Alpha Centauri is by having a magnetic sail ship follow the Sun's south polar jet. You get *continuous* acceleration almost for free (you must still adjust for the bends in the magnetic field lines, since the ship's inertia is much higher than an ion's). You won't go over the jet's speed, and so won't reach Alpha Centauri in 50 years, but 150 years is not dramatically different, and you aren't limited to 1 kg; the ship can be of any mass.

Of course, this is predicated on the "open" magnetic field lines of the Sun reaching all the way to Alpha Centauri. Then, the jet (aka "flux ropes") can't help but follow; plasma behaves completely different from gas, which wouldn't have followed a great distance.


Charlie@171: Imagine for a moment that humans started reproducing *exactly*. Moreover, that the neurons in the growing brain of an infant mysteriously started establishing connections in exactly the same way (they don't; which is why not even twins or clones can have the same brains). Do you think the loss of variation would be: (A) beneficial, (B) disastrous, (C) not have much of an effect.

I bet on B.


May I throw in another consideration into this discussion?

There's another problem with our thinking about star *SHIPS* and that's the assumption of the medium that any putative star, um, thingy, is traveling through.

It's traveling through a vacuum, right? Space is a vacuum, we've been told again and again and again.

Well, not at any appreciable interstellar travel speeds it ain't. 1 H2 molecule per CC is a pretty damn good vacuum, but not at ten percent of lightspeed. At that sort of speed, 1 molecule of *anything* is, instead, a pretty good gamma ray producer, if not also a pretty damn good ablator.

"OK," says my handwaving evil twin, "what if we have some groovy strong magnetic fields to shield us." Says I to my evil twin, well, gee, that takes care of charged particles, but doesn't do squat for the neutral ones, much less any dust grains.

I don't want to even *think* about what a dust grain does to something traveling at ten percent C.

And that's just at my arbitraily selected ten percent. If one is talking about higher speeds -- 70, 80, 90, 99 percent of light -- well, I don't know how you even manage to keep from being boiled away at those speeds from the impacts.

Charlie was right in his initial post -- interstellar flight will *have* to be *SLOW.*


Tim@176: The situation isn't quite this dire--depending on *where* you travel. If you follow a magnetic field-aligned plasma filament (aka "flux rope") a la auroral currents, then neutral matter will accumulate in "hotspot" disks along the way (where the Buneman plasma instability occurs and a double layer forms), with hardly any in between. And what neutrals are in between would be a part of the plasma flow; the ionization degree doesn't have to be particularly high to involve the neutral fraction in the flow. So, your task is mainly to detect the hotspots up ahead, which shouldn't be a problem. Plasma is a very good conductor, so the hotspots will be far between.

As to the magnetic field to deflect charged particles, it doesn't have to particularly strong.


Heh, for all we know, the Sun may be one such hotspot, and Alpha Centauri another, with none in between. :)


Greg@172: there actually were experimental powered flights made with steam engines, and they worked. But that was around the time that petrol engines were getting light and powerful, so nobody pursued it.


Leonid@178 - you keep talking about these hypothetical flux tubes from the sun's poles. What evidence do you have that they do not simply curve around to the other pole as one might expect from the forces?


Alex@180: Simple. Even in the plane of the ecliptic, the "slow" solar wind becomes superalfvenic (the speed higher than the Alfven speed), and therefore the magnetic field lines can never contract back to the Sun. Never mind the "fast" solar wind in the polar regions.

You may have pictured the Sun's magnetic field as that of a simple N-S magnet. But this is not the case. The Sun's magnetic field is substantially more complex.


@Tim Kyger #176

As far as single particles are concerned, speed doesn't matter: most of the background radiation is extremely high energy with sizable gamma factors and it doesn't matter a bit if you are parked or go at 0.5 c. To magnetically deflect those beasties, mostly protons and alpha particles, you require about 100 Teslameters - well beyond any compact superconductor configuration, so you would need gigantic coils with several Hiroshimas worth of field energy which you will have to prevent from quenching for centuries or millennia ...

The good new is, that some meters of stuff will stop them. Hydrogen works best from a mass perspective - so you might get away with less than one ton per square meter (more if you want to use methane, ammonium, PET or water). Incidently, Hydrogen also makes the best reaction mass for many propulsion schemes, so if you have a reaction or fusion drive with high enough thrust so that you can do a seizable portion of your breaking within the safety of the targets heliosphere, you radiation shield is essentially free.

Dust grains at 0.1c are another matter, but I guess those could be taken care of by a thin, boom mounted aluminium plate some hundred meters in front of the ship which should spread out their 1 to 100 kg TNT yield to managable levels.


Amputee dwarf women.

That's what will get us to the stars. We don't need men, just a fridge. We don't need useless limbs to feed and propel.

Amputee dwarf women.


Ignatius@182: I didn't even consider the cosmic ray stuff. For those, the ship's speed indeed doesn't matter. I thought they were talking about the (relatively) slow-moving plasma.


And, a word of advice: Dump solar sails. They are inferior to magnetic sails in so many ways it's almost obscene.


J@168: Nah, just AI. I think AGI will require at least one further trick. When I meet a computer that wants to tell me about how it found Jesus, and convinces me that it is sincere, then I'll believe we have AGI. Don't get me wrong, I'll be face-palming so hard it hurts, but behind my palm I'll be smiling.

I'm serious about calling these things tricks. I think once we work out how to do AGI it will seem pretty straightforward. It might even be so obviously lacking in magic spark that we will fail to acknowledge it as our moral equal, for fear of what this would imply about our own nature. It won't be a matter of blindly copying the human blueprint, and... ta-da... human in a box that we still don't fully understand and can't tinker with. Nor will it be a matter of building a machine so that we deliberately don't understand how it works (see the failure of "neural networks" to live up to their hype). It'll be bread-and-butter maths, statistics, computer science. The sciences, each straining in its own direction, have hitherto harmed us little; but some day the piecing together of dissociated knowledge will open up such terrifying vistas of reality, and of our frightful position therein, that we shall either go mad from the revelation or flee from the light into the peace and safety of a new dark age.


@Leonid #185
Wiki excerpt"
Although solar wind particles have rest mass and photons do not, sunlight has thousands of times more momentum than the solar wind. Therefore, a magnetic sail must deflect a proportionally larger area of the solar wind than a comparable solar sail to generate the same amount of thrust. However it need not be as massive as a solar sail, because the solar wind is deflected by a magnetic field instead of a large physical sail. Conventional materials for solar sails weigh around 7 grams per square meter, giving a thrust of 1e-5 N/m2 at 1AU. This gives a mass/thrust ratio of at least 700 kg/N, similar to a magnetic sail, neglecting other structural components.
The solar and magnetic sails have a thrust that falls off as the square of the distance from the sun"

In brief: Magnetic and solar both have advantages. Solar sail is limited only by material science whereas magnetic sails require some work.

I think using methods in a hybrid system could use the benefits of both including the ability of solar sails to tack.


edit Magnetic sails can tack as well and becomes even more useful around the gas giants with large magnetospheres. But I wouldn't discount solar sails yet since in combination with magnetic sails they can add to the maneuverability of the craft.



SoV @170: Tsar Bomba was designed to go to 100Mt; for the actual test they replaced a chunk of the U238 booster with lead because (a) they didn't want to melt the Tu-95 from which the bomb was dropped, and (b) they didn't want the fireball to touch the ground and spray highly radioactive fallout all over the USSR. The projected yield without the booster layer -- 50Mt -- was indeed a close match for the design, so there's zero reason to suppose that Tsar Bomba wasn't a viable 100Mt design.

If you look at what I've posted, I think you'll agree that I'm already aware of all of this, and have been since before I asked. This is a side point; even if a 100 megaton bomb has never been dated, the energy yeild would be the same as ten 10 megaton bombs, and makes the intended point just as well. IOW, I just want to know what the biggest barbaric yawp that has ever been made by a human being :-) However:

As for the cost of energy from nuclear weapons: we know the US nuclear weapons program cost on the order of $2Tn to date. (Figure from a back issue of SciAm; can't be bothered to go digging for more accurate data online right now -- got to go out.) From this, and the number of weapons produced (about 30,000), and the average yield (around 50Kt?), and the stockpile of Pu239 left over (200 tons, or thereabouts?), we can guestimate the energy output of $2Tn in joules (i.e. watts/second).

This is irrelevant to the issue. The cost under consideration is not the cost of any old 10^17 J, but the cost of a usable 10^17 J. Suppose you detonate this 100 megaton bomb under water (or ten 10 megaton bombs). 10^3 J will raise the temperature of one kilogram of water one degree Centigrade; so this much energy would raise the temperature of 10^14 kg 1 degree C. That's not really all that much water, incidentally, just 10^13 cubic meters. So how much usuable energy can you extract from this one degree difference of heat? If the water was initially at 25 degrees C, then elementary thermodynamics says the very best efficiency you can hope for is (26-25)/(273+26), or about one third of a percent.

The lesson here is that it's not the cost of energy that's the defining parameter here, but the cost of usable energy. It's no good simply to say that this bomb cost X and it released an amount Y, and so the unit cost for energy is Y/X. Not when what comes out is heat, gamma rays and fission products. That's why the black box model of the drive is so nice; it matters not what goes on inside, whether it's absolutely mundane stuff, or whether it's a reactionless motor that can accelerate anything within it's field of influence at hundreds or thousands of g's. No matter what, you've still got to come up with the energy necessary to power it[1]. And short of another black box breakthrough, you're not going to be able to suck all the energy out of a nuclear explosion and then convert it into usable form for essentially zero cost.

[1]Another hypothetical: think of using "free" energy from the Sun to power a laser that will run a, say, 100 kg light sail design up to .866 c. Sure, that's going to take more than 10^19 J, but hey, sunlight, like butterflies and the laughter of little girls is free, right? Now ask yourself how much it costs to produce a one kilowatt beam of laser light from a solar powered source. I'm guessing that you're going to say it's going to cost more than $0.05 per hour of run time. Iow, my $0.05/kWh pricing is most decidedly not pessimistic, nor was it intended to be.


Dalani@187: The major advantage of magnetic sail is its superior maneuverability, far beyond any other means of propulsion. A solar sail is pushed forward, a magnetic sail is deflected sidewise. If you have more than one magnetic sail, you can set up them up in many different adjustable configurations; such a ship can literally dance in place.

Moreover, a magnetic sail is a coil, not a surface. It can shrink and expand automatically, depending on the flux. But if it catches a magnetic field-aligned filament, held in shape by the electromagnetic pinch, then the flux doesn't fall with the distance from the Sun.


@Leonid #190

Indeed, you are referring to the Mini-magnetospheric plasma propulsion method, not quite a classic magnetic sail-it does need to carry a quantity of gas, but does hold great potential. But think of the weight of those coils! like I was saying, magnetic sails need more work: eg cooling methods, weight, not to mention superconductivity materials.


Dalani@191: Well, the mini-magnetospheric propulsion is really a variant of magnetic sail. The only difference is how the magnetic field is formed, but the dynamics of propulsion are identical. Personally, I think the coil method is more reliable, because the mini-magnetospheric method's plasma can be simply blown away in certain situations, joining the solar wind. The coil too can be paid out and pulled in automatically, preserving the thrust, but to a lesser extent (limited by the total length).



All your talk of energy cost has me thinking about dusting off my yellowed copy of Pier Anthony's Cluster Series: the stories concerned various scenarios of the prohibitive cost of energy to transmit physical matter, even energy wars between galaxies and would you believe he had one scenario pitting Andromeda against the our Milkyway: to rob the energy of our galaxy. -all this decades before astronomers recently confirmed Andromeda had eaten smaller galaxies and could threaten our own.

As for free solar energy : most abundant energy sources are free (or very cheap) until it seems its not enough for the purpose. A persons legs can propel a bicycle easily but to get an airplane to fly requires an engine with higher energy needs (and higher operating temperatures). the glass ceiling here is obviously E=mc2. But greater energy can achieved with less through evolving greater efficiencies. Doing more with less was not a factor in the Victorian era..steam engines got bigger and smokier.. That thinking was Piers and your assumptions about energy use for propulsion. Who's to say, that with methods other than the brute force approach, we cannot achieve effortless transport with nothing more than the low energy medium of solar wind or interstellar particles.


@ 193: "Doing more with less was not a factor in the Victorian era..steam engines got bigger and smokier.. "


Towards the end of that period, people were getting VERY concerned about the lack of efficiency of steam-engines, both stationary and moving.
Coal COST, and moving it about cost as well.
Basically there were three approaches to this problem.
1. Compounding
Used principally for marine and fixed engines, where the size was large, and roratinal speeds relatively low.
Some marine engines were triple-expansion, and eben one or two cases of quadruple. The local Waterworks here replaced their double-expansion compunds with triple, as late as 1927 - and they only ceased working when they moved the pumping station, about 15 years ago .....
Was used on railways, especially in France & Germany, but not here, as the slightly tighter loading-gauge made big LP cylinders impractical.
2. Superheating
Especially for railway use.
Dropped the coal consumption by something like 50% in the best cases, whilst boiling more water as well.
Was followed-through by changing the valve-events and draughting of locomotives.
3. Turbines.
Look up Charles Parsons, or "turbinia".
Of course turbines could ALSO be multiple-expansion, especially if used in large ships.
The development processes for turbines are still with us, of course.


Paul @186: The sciences, each straining in its own direction, have hitherto harmed us little; but some day the piecing together of dissociated knowledge will open up such terrifying vistas of reality, and of our frightful position therein, that we shall either go mad from the revelation or flee from the light into the peace and safety of a new dark age.

Are you channeling H. P. Lovecraft intentionally, or do you think AGI is intrinsically likely to make people start going "Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn!" on sight?

SoV @189: you're misinterpreting me: given what we know about critical masses, and about availability of plutonium, we can guess reasonably accurately about how much Pu239 and HEU the US government produced with that $2Tn. And then at how much energy we could extract from it using available tech (e.g. MOX fuel for fission reactors).

(It turns out that burning Pu239 and U235 in reactors is rather inefficient compared to the lightly enriched uranium used in commercial PWRs, which is why that's what the commercial sector does. In fact, that whopping great 100Mt H-bomb with it's 10^17 joules gives us roughly the same energy output of a 1Gw PWR running for 30 years. Who knew?)


Greg Tingey @ 194
1. Compounding
Used principally for marine and fixed engines, where the size was large, and roratinal speeds relatively low.
Was used on railways, especially in France & Germany, but not here, as the slightly tighter loading-gauge made big LP cylinders impractical.

By "here" do you mean Britain? 'Cos if you do, I'm sure that my father's book of British steam locomotives included examples of compounding.

Of cours, that may have been later than Victorian.

J Homes.


@ 196
YES Britain.
Webb's LNWR compounds were hopeless, and the Smith/Johnson ones on the Midland did well - but they had 200/210 lb boiler-preesures when everone else was using 160 ...
They were hopeless when really heavy loads arrived, or your journeys had frequent stops - partly because of faction-fights inside the MR in 1910-12 - they WERE going to build a large (for the time) 4-6-0, but it got cancelled, and they stuck with little 4-4-0's - until 1926 (!)
By which time Gresley on the LNER was building decent pacifics - and the railway world was changed for ever.

NOTE (Sir) H. N. Gresley designer of:
The largest steam loco in Britain (Class U-1, 2-8-8-2 Garratt type) The largest passenger loco in Britain (class P-2) the fastest steam loco on the planet (class A-4)

Re. previous comments - even Watt's invention of (then) "high-pressure" ( ~ 8psi) steam power in the 1780's was a response to increasing costs and inefficency.

SoV @189: you're misinterpreting me: given what we know about critical masses, and about availability of plutonium, we can guess reasonably accurately about how much Pu239 and HEU the US government produced with that $2Tn. And then at how much energy we could extract from it using available tech (e.g. MOX fuel for fission reactors).

Charlie, you've priced just the energy part. How much does it cost to extract the usable energy? You seem to be assuming that the cost there is zero. So how are you using your nuclear stockpile to power your starship? Are you converting it to beamed lasers/masers? Is it directly powered by electricity? What?


@Greg #194 No it turns out I'm 100% correct: The progression of steam engine development- especially steam locomotives - as been towards greater power output with bigger engines not smaller more efficient engines to save coal cost. Coal was always plentyfull to this day. The first Stephenson Rocket had a fraction of the power of much bigger engines that came later. Check any specification of engines built from 1800 onwards you'll see these facts plainly. Behemoths like the Canadian Pacific Hudsons and Sir Nigel Gresleys had thousands of Horsepower , greater speed and range and yes used a sh*itload more coal than the first Stephensons and used compounding-all in the name of power and profit. It wasn't about saving costs* but hauling more freight and passengers further, faster using even more coal. BTW locomotives did not get less smoky in fact this was such problem it drove innovations of diesels and electrification onto the railways in Europe.


@Leonid Magnetic sails: ARE THEY SAFE?
The effects of strong magnetic fields on the human body is not something I want to deal with at this late hour. Show me the money.


Charlie@195: Let me try it again. Bituminous coal has an energy content of about 3.3x10^7 J/kg. Since a kilowatt-hour is 3.6x10^6 J, I'll approximate the energy in one kilogram of coal as about 10 kWh. Now, bituminous coal was selling for about $51/ton, that is $51 for 2,000 pounds, call it $0.06/kg. So 1 kg/10 kWh*$0.06/kg = $0.006/kWh. That is, one kWh worth of energy from coal costs just 6/10 of a cent. Now, looking up my rates, it runs about $0.10/kWh. That's residential, of course, but even the general rate for large concerns is still about $0.05/kWh.

So where did all the extra charges come from? Our utility is a traditional monopoly concern that is regulated by the government, so there's no greedy capitalist blood-suckers to raise rates they way they've been raised in other parts of the country when utilities were deregulated. The cost per kWh of electricity generated by nuke plants is even higher, btw. Using fission to generate electricity won't cut costs(at least, not in the here and now.)

Now, back to your nuclear stockpile. How do you plan to use the energy content of your assorted weaponry, and how much is it going to cost to put that content into the form you desire? You could say that all you want is a really big bang, of course, with all the appropriate references to an Orion-type of propulsion system. But in that case, you can not simply say that the 10^17 J/kg is all that you need to get to .866 c; now the energy requirements are much, much higher. In fact, since the maximum velocity of your fission fragments is - at the very best - about .03 c, the rocket equation says your mass ratio is about e^28, or about 3 trillion to one.

I hope I've explained this better on the second or third go-around.


Dalani@200: Magnetic field doesn't have to be strong everywhere. As to a relatively small magnetic field, I would argue that we *want* it. Consider that we were born--and all our predecessors, not limited to humans--on a planet with a significant magnetic field. We may actually need it to be healthy.



Try thinking about the figures that were always quoted at the time...
Coal consumption - lbs/mile
Coal use efficiency lbs/DBHP/hr
and sometimes evaporative rates as well.

So that, by the late 1930's the better locomotives were travelling at 65 mph average speeds, and burning less than 35lbs of coal per mile, whilst hauling a 350 ton train.
As opposed to unsuperheated locos in the 1890's doing 40mph averages, with 150 ton trains, and burning 50 - 70 lbs of coal per mile.



What's more this had all happened before - in the 1790 when Watt/Boulton stationary engines used the same amount of coal, or slightly less than their predecessors, but put out a LOT more power.


SoV: it doesn't matter how we use the energy from the Pu -- for the sake of argument: stick it in MOX reactors, maximize energy extraction (around 35-40% -- Carnot applies) and figure out how to use it in a 100% effective (handwavium, I know) laser to power a light sail. It still gives us a rough benchmark for the money-to-power ratio we can get from nukes (via a weapons-oriented program, admittedly). The interesting factoid is that the total world nuclear weapons stockpile is only enough to power our regular electricity cycle for on the order of a year (give or take a factor of 2-5). For starships, it's pathetic -- unless we stick to low (


I think we'll be able to make AGI anything we want them to be. Perfectly nice fellows, if we want. We'll be able to lay open their brains, see exactly how they think, and tinker with it. Shortly thereafter, we'll be able to do this to ourselves. Complete freedom. There's an element of existential horror to this scenario -- not just because bad guys might get hold of these options, but because of what we might find out about ourselves and what we might willingly do to ourselves. Hence the Lovecraft quote.



SoV: it doesn't matter how we use the energy from the Pu -- for the sake of argument: stick it in MOX reactors, maximize energy extraction (around 35-40% -- Carnot applies) and figure out how to use it in a 100% effective (handwavium, I know) laser to power a light sail. It still gives us a rough benchmark for the money-to-power ratio we can get from nukes (via a weapons-oriented program, admittedly).

This is wrong on two counts, one wildly so, the other mebbe "right" if you mean "within two or three orders of magnitude".

Let's do a quick-and-dirty laser powered light sail: let's say it's a square kilometer of sail, and sail and payload together mass 10 kg. Further, we'll assume a laser that hits just the sail, and has an insolation of 10^4 W/m^2. Then the total power output of the laser is on the order of 10^10 W, or 10^9 W/kg. Double this (assume the sail is perfectly reflective) and divide by c, 3x10^8 m/sec, to get the total momentum transfer per unit time; about 6.7 m/sec. Rounding up, we can say that this setup imparts an acceleration of roughly 10 m/sec^2, so that over the course of one second, velocity increases by roughly 10 m/sec. That's pretty good for this sort of setup, but it's still an overall very poor showing. Given a power of 10^9 W/kg, how fast would it be theoretically possible for this probe to go? Well, that's just the old kinetic energy formula, E=0.5 mv^2; solving for v and then plugging in 10^9 J and 1 kg for E and M, we see that the theoretical upper limit is a bit better than 4x10^4 m/sec! Iow, in one system, you get a total velocity change of 10 m/sec, in the other, 4x10^4 m/sec. That's off by a factor of 4,000. So, yes, it makes a great deal of difference how we use the energy from the Pu.[1] I'm saying that it takes a theoretical minimum of 10^17 J to get a kilogram moving at 0.866 c. That's basic physics. But depending on how you accelerate this kilogram of material, the energy requirements can be much, much, much higher. That's why I asked just how you planned to use your supply of nuclear material.

On the other point: I've already posted an analysis that says the energy in coal should theoretically cost about $0.006/kWh. I, however, am charged about $0.1/kWh. Even taking into account various thermodynamic and mechanical inefficiencies, my price at the meter is still five times higher than it "should" be. Why is that? It's certainly not due to evil capitalist pig dogs. The same analysis applies to nuclear of course, only worse (that might be the result of other factors here in the U.S. as opposed to say, France.) Any thoughts on this one? Note that here we're talking of factors of ten or so, probably less, so that as far as a BOTEC goes, it's not really wrong, if that's one of the points you were getting at above.



SoV: it doesn't matter how we use the energy from the Pu -- for the sake of argument: stick it in MOX reactors, maximize energy extraction (around 35-40% -- Carnot applies) and figure out how to use it in a 100% effective (handwavium, I know) laser to power a light sail. It still gives us a rough benchmark for the money-to-power ratio we can get from nukes (via a weapons-oriented program, admittedly).

This is wrong on two counts, one wildly so, the other mebbe "right" if you mean "within two or three orders of magnitude".

Let's do a quick-and-dirty laser powered light sail: let's say it's a square kilometer of sail, and sail and payload together mass 10 kg. Further, we'll assume a laser that hits just the sail, and has an insolation of 10^4 W/m^2. Then the total power output of the laser is on the order of 10^10 W, or 10^9 W/kg. Double this (assume the sail is perfectly reflective) and divide by c, 3x10^8 m/sec, to get the total momentum transfer per unit time; about 6.7 m/sec. Rounding up, we can say that this setup imparts an acceleration of roughly 10 m/sec^2, so that over the course of one second, velocity increases by roughly 10 m/sec. That's pretty good for this sort of setup, but it's still an overall very poor showing. Given a power of 10^9 W/kg, how fast would it be theoretically possible for this probe to go? Well, that's just the old kinetic energy formula, E=0.5 mv^2; solving for v and then plugging in 10^9 J and 1 kg for E and M, we see that the theoretical upper limit is a bit better than 4x10^4 m/sec! Iow, in one system, you get a total velocity change of 10 m/sec, in the other, 4x10^4 m/sec. That's off by a factor of 4,000. So, yes, it makes a great deal of difference how we use the energy from the Pu.[1] I'm saying that it takes a theoretical minimum of 10^17 J to get a kilogram moving at 0.866 c. That's basic physics. But depending on how you accelerate this kilogram of material, the energy requirements can be much, much, much higher. That's why I asked just how you planned to use your supply of nuclear material.

On the other point: I've already posted an analysis that says the energy in coal should theoretically cost about $0.006/kWh. I, however, am charged about $0.1/kWh. Even taking into account various thermodynamic and mechanical inefficiencies, my price at the meter is still five times higher than it "should" be. Why is that? It's certainly not due to evil capitalist pig dogs. The same analysis applies to nuclear of course, only worse (that might be the result of other factors here in the U.S. as opposed to say, France.) Any thoughts on this one? Note that here we're talking of factors of ten or so, probably less, so that as far as a BOTEC goes, it's not really wrong, if that's one of the points you were getting at above.

[1]This sort of analysis is common in theoretical rocketry. For example, it turns out that the optimum exhaust velocity of your rocket should be about 2/3 of your total mission delta-v. This seems counter intuitive to some people; they think the higher the Isp the better. But this turns out not to be the case when analyzed properly; in fact, high Isp is actually quite wasteful.


@203: Yes the engines put out more power..why? because that was the demand ...period .There was no shortage of coal


dalani@203: do you understand what Greg means by the term "efficiency"? It's power per unit fuel. You appear to think that small and efficient are the same thing.

It might shock you to learn that the most efficient piston engines are the giant diesels that power ships, burning hundreds of litres an hour.


@ 209
Think about the diesels that supplanted the A-4's out of King's Cross.
A "Deltic" had a fuel consumption of about 3mpg.
BUT hauling a train of between 300 and 400 tons, plus its own weight of 100 tons.
Meanwhile a family car of the period would do 30 mpg on a good day, weighing under 2 tons, even when very fully laden
Which is the more efficient?
Paricularly as the Deltics were easily 100-mph capable with that load.



YOU ARE OVERLOOKING THE BROADER PICTURE OF SOCIETAL PRESSURES ON THE DEVELOPMENT OF TECHNOLOGIES: yes of course a 3000hp diesel railway engine uses less fuel than a prius if you look at ton/mile/gallon. But the prime motivation of developing the CP Hudson steamers was not to use less coal to do more but rather that efficiency came about as a byproduct of large engineering budgets aimed to fulfill the market demand for hauling more ton per mile faster.

If efficiency was the only motive than either the Stephenson Rockets would have have evolved into MANY machines with the same power output OR all those Stephenson Rockets would have been replaced by FEWER Hudsons doing the same work. And we would all be taking trains to work, cars would have never existed, and the main highlighted specification of 19th century locomotives would not have been HORSEPOWER or TOP SPEED but fuel cost per mile. Your equations are only valid for engineering not for uncovering cultural premises.

By isolating your focus to a few equations of thermodynamic efficiency you are overlooking the REAL societal forces at work.


@211: Dalani, getting back to base topic here (starships), I'd say that efficiency discussions here are relevant. Barring some discovery of FTL travel that consigns current cosmology to the history books, we're going to need a huge amount of energy to move a tiny amount of mass on stellar distances, to the point where Charlie (and I, and many others) are talking about sending simulations and records of humans and nanotechnology to build new bodies at the other end, on voyages that will take centuries to millenia.

That's all about efficiency at all costs. As I noted in a previous post, you've got to be willing to die to fly this way. It's a different cultural milieu than the family car.

@Charlie: reopening an old issue... If we have a beam-riding starship, it doesn't particularly matter whether the starship is tracking the beam or the beam generator is tracking the starship: aiming is one hell of a technical problem. Assuming a 100 km wide sail at 1 light year, to stay on the sail, the beam has to stay pointed accurately within 6e-10 degrees (which is on order of holding a one kilometer object in position to 10 nanometers). A one degree twitch of the generator's aim will move the beam 1.6e11 km laterally at one light year. It's hard to get a sail re-trimmed to tack laterally that fast.

Unless we have a beam that can correct its own aim mid-flight, I'd suggest accuracy is an even bigger problem than efficiency for beam-based propulsion.


On an ENGINEERING construct, owned by commercial companies?

What planet are you on?


One issue: if we have the kind of GAI that this starwisp probe needs, then much of the starwisp design becomes irrelevant.

A GAI doesn't ever have to die, so the most likely *sponsor* for this space adventure will be another GAI back on Earth.

The mission can take thousands of years if need be, which removes a lot of the energy and mass constraints of the whole project.


Now some major hand-waving just for fun...

Judging by this conversation, we're going to be trapped in our solar system for a long, long time. Our industries and bodies could be festering here for millenia.

In that much time we could have converted our whole solar system into an industrial base. Even without true self-replicators, we could expand explosively with finite, heirarchical meta-manufacturing...

So: consider a telescope/phased array with an aperture the size of our solar system, and a frequency range from microwave to gamma.

Such an array could burn *shapes* into the surfaces of bodies far outside itself, and the first shapes you'd want to burn would be mirrors, to do undercuts.
(Also reference, Laser Focused Atomic Deposition: [http://cnst.nist.gov/epg/Pubs/pdf/epg717.pdf )

Could its manipulation reach other star systems? I doubt it, but the array would be highly aware of every interstellar body on the way to those systems, and could make a drastic modifications to its interstellar environment.

It could ablate rocks for light-years around into yet more mirrors, and daisy-chain its control to expand yet further outward.

So I'll up the ante: the picture isn't ships or even probes. The big picture is a 100,000 AU cluster intelligence slowly expanding tentacles of influence into interestellar space, and only touching other stars when it's Damn Well Ready.


@212 I agree with you 100%: This whole digression is simply an assertion I made that Victorian age did not have the SAME market demands we now have or will have. The Toyota Prius could not have been successful in the Victorian age (yes electric cars existed then)-range and power mattered more than emissions-thus gasoline became the fuel of choice.

@Greg societal forces .as in large sums of money and large resources committed towards certain developments. In the medieval age resources were committed towards developing and making pulley systems to help build cathedral. Pick any time period and you will see different motivating factors behind technological developments (or other developments) . Im not the only one saying these things re-read post #13.
But I leave you with this :
consider glass made with 1000F furnaces in some Victorian cauldron glass mill Vs diatomic algea quietly secreting glass shells in the cold waters of the artic.


@214: John, I think there's a plot there for a novel alien invasion scheme, if the alien laser array is aimed towards the solar system to make communicaton easier. It's a flashy idea anyway.


I'm trying to figure out how to transfer momentum from the solar system to a departing starship. You can't steer photons, but if there's a way to cheaply create a high speed mass that could be steered without loss of mass for reactions, it would be useful. Something that's highly charged (i.e., a large body of plasm) might be steerable, especially if it is traveling in a medium that also has charged particles.

While I hate mentioning plasma vortices because of the crop circle associations, I wonder if it's possible to generate a very stable plasma vortex and use it to transfer momentum to a starcraft. This would probably work if it was possible to encase a small vortex shepherd robot spacecraft inside a large mass of plasma that's rapidly accelerated like a relativistic smoke ring. The shepherd helps keeps the vortex together and steers the darn thing, and then the vortex impacts on the starship's magsail and dumps its momentum into the spacecraft (and possibly replenishes a shield field, depending how the magsail works.

Is this the equivalent of blowing a supersonic smoke-ring, or is it even theoretically possible?

I'm trying to figure out how to transfer momentum from the solar system to a departing starship. You can't steer photons, but if there's a way to cheaply create a high speed mass that could be steered without loss of mass for reactions, it would be useful. Something that's highly charged (i.e., a large body of plasm) might be steerable, especially if it is traveling in a medium that also has charged particles.

There have been all sorts of proposals for just this sort of thing; you might want to google "pellet stream propulsion". The seminal article seems to be by a Clifford Singer, "Interstellar propulsion using a pellet stream for momentum transfer", which is described in the abstract as:

A pellet-stream concept for interstellar propulsion is described. Small pellets are accelerated in the solar system and accurately guided to an interstellar probe where they are intercepted and transfer momentum. This propulsion system appears to offer orders-of-magnitude improvements in terms of engineering simplicity and power requirements over any other known feasible system for transport over interstellar distance in a time comparable to a human lifespan.

Here's another reference. The basic idea is very simple and relies on just two ideas, conservation of momentum, and conservation of energy; ideally, you want all the energy to go into the object to be moved, and very little into the ancillary mass. Physically speaking, the way to do this is to make the moving stuff much smaller than the mass that is doing the accelerating, with the ratio of the KE(and thus the velocity) differences going as the square of the mass ratio. Since pellets on massing on the order of a kilogram will be orders of magnitude smaller than the accelerator, massing on the order of several thousand kilograms or more, virtually all of the energy ends up in the pellets (or ship, for that matter.)

This is actually very old science, but for some reason doesn't seem to widely known or accepted. I guess laser-launched light sails just have better optics.


@218: Thanks SoV: I've seen this in Orion's Arm beamrider network. Your description makes more sense than theirs, because they talk about steerable micropellets, and I was having trouble figuring out how you steer something that small. Presumably a micropellet would have micro delta-V, but who knows. Maybe you shoot off a huge mass of micropellets that are each three-quarters anti-matter. That would work. Sure...

I was considering plasma, because I know that ocean currents can generate and carry vortices for quite a distance. Presumably solar plasma carries vortices too, and I was wondering if there was a way to stabilize and aim such a vortex to turn it into a way to deliver momentum.



RE: stable plasma vortex

Look up "Spheromak"

It's exactly what you're describing. They can even be fired out of a gun-like device. I doubt it could ever be useful at extremely long range though.



Let's try that again:



Thanks Hylas, that's what I was thinking of. You know, compared to, say, an antimatter drive or even a fusion torch, this sounds almost doable. You really need to do is postulate that some sort of device/robot spaceship can be embedded in the bubble to keep it intact until it reaches its target.

Compared to maintaining kilos of antimatter for centuries at a time (and releasing the AM on cue to move the ship), keeping a bubble alive would be easy.

Moreover, if we postulate that tokamak-style fusion will actually work as a reasonable power source, this is much the same technology. Instead of creating a high-temperature stable containment field, you're creating a light weight, ambient temperature, moving containment field. That's almost, sort of, halfway reasonable as a star drive.



I was considering plasma, because I know that ocean currents can generate and carry vortices for quite a distance. Presumably solar plasma carries vortices too, and I was wondering if there was a way to stabilize and aim such a vortex to turn it into a way to deliver momentum.

There's a lot of different ways to work this scheme, but if you insist on plasma appearing somewhere, one variant has a shipboard laser zapping the pellet into it's component electrons and nuclei. Another variant just squirts out a little dust or gas and lets the pellet run into it at a relative velocity of several dozen or hundred k.p.s. Presto! Instant plasma. As a bonus, you can also tap it for energy even as it's accelerating the ship, that whole current-carrying conductor thing. One nice consequence of using current loops to interact with the pellet/plasma rather than a physical plate is that they can be extremely large for relatively little mass, since mass will scale almost linearly with diameter rather than as the square of the diameter as it would with a light sail.


@324: whichever works. So far as I can tell, the big downside to steerable drive beams is making sure that the beam components don't shed all their delta-V steering their way into the target starship, and don't get shredded by the interstellar medium.

To me, that's the problem with pellets. I'm pitching the idea of using some form of plasma vortex (or if you want, a smaller, fast-moving magsail that contains a large amount of plasma confined by the sail) as the unit of momentum transfer.

I'm sure there are all sorts of problems with it, but there are problems with all interstellar drives.


Somewhat OT, but wouldn't the high powered laser/maser technology be really good for just powering spaceships (e.g. VASIMR powerplants) around the solar system. It's almost retro - I believe Dan Dare's spacefleet ships were powered by some sort of power beam.


I'm surprised that no-one has commented that there is no limit on the subjective speed of travel.

We think of computers as devices to make things happen faster, and extrapolate a singularity for the time when they are capable of running an intelligence orders of magnitude faster and/or sophisticated than meat brains.

However, there's another possibility. Given machine intelligence (or uploaded intelligence), run it on hardware that can slow its hardware clock down by orders of magnitude. Leave the solar system at mere Voyager-type speeds, with an absolute travel time measured in tens of thousands of years. But slow the CPU clock down ten-thousand-fold, and subjectively it's a journey of a few years at subjective speed ~c. Or a hundred-thousand-fold, and it's a few months subjective at ~10c.

On arrival, speed up the clock again. (Practically, you'd probably decamp from a micro-power slow CPU into a more energy-hungry faster one running on solar power).

The problem then reduces to whether it's possible to design hardware that can survive tens or hundreds of thousands of years in interstellar space. Cosmic ray damage would be the big enemy, a lot of redundant circuit elements ought to solve it.

And, of course, there's a bigger problem, whether the virtual occupants of this subjectively superluminal starship can re-boot a civilisation at the far end. Although perhaps that wouldn't be essential. Maybe a virtual community would be happy merely to see a few other star systems and then return to find out what has become of Earth several million years after departing, even if the most likely end-point was death after a human-equivalent lifetime of sight-seeing.


Nigel: I'm surprised that no-one has commented that there is no limit on the subjective speed of travel.

Haven't read "Saturn's Children", have we?


Found a new drive system, from Peter Watt's Blind Sight: using a long-range atomic/quantum teleportation system to teleport antimatter particles into something like a from the AM manufacturing facility near the sun. Entangling information was sent via radio, and the antimatter was fed into something like a Bussard ram.

It's different, anyway. This is the same story where he turned people into vampires temporarily so that they could ship out as "undead" in suspension crypts, and be revived at the other end. I was reading after the finding out about the mess that happened to Peter...



Actually, black holes should be relatively straightforward to contain, as long as you have plenty of electrons to feed them.



Given the population demographics in the developed world our hypothetical future state may never have the impetuous to reach the stars. It may well be a rather small population (say 200 years of decline met with life extension) who spend a good deal of time in interactive V.R. worlds that makes up humanity

We may find it rather difficult to recruit someone for a one way mission thats likely far duller than what they have already experienced.


Talking about interstellar travel before we have even expanded into the solar system... That's sort of like a group of hunter-gatherers living in huts on an island with a couple of rafts ... and talking about travelling to a continent thousands of kilometers away, when they haven't even settled the other small islands in their island chain, just across the water.

The solar system, right in our back yard, itself is *enormous*. It's like the Hitchhiker's quote ("Space is big. Really big. You just won't believe how vastly, hugely, mind-bogglingly big it is.") while substituting in the term "solar system".

And so are the resources that it presents us. It has enough to support a population of humans that -- literally -- could be anywhere from millions to billions of times larger than our current piddly few billion people. The total economy and control of energy resources, will be multiplied by a similar factor. The sun puts out 3.84e26 watts, compared to the 5e14 watts used by all of humanity currently. Exploration and settlement of the solar system, and then developing the civilization from that, will occupy the next several thousand years, will be an epic tale.

After that, with a Kardashev Level II civilization, the cost and energy required for interstellar travel will be nothing more than a minor problem. They will be able to it the way they *want* to do (using real humans, for example), not just the cheapest and expedient


I remember doing froth-at-the-mouth rants professionally, even, something on the order or "a spacecraft is NOT an a water surface vessel. A spacecraft is NOT a submarine. A spacecraft is NOT an atmospheric vehicle. A spacecraft is not a land vehicle. The lowest orbit that a satellite won't fall out of in two weeks is more than 100 kilometers above the planet's surface. Objects in orbit are not -stationary-, they're in orbit, and unless something else is along the same orbit (and even then, without orbit adjust capability, due to inhomogeneities in the planetary gravitational field and perturbations from the solar wind, atmospheric drag, the mass of Luna, etc....), the positions of the two objects -change- with respect to one another as a function of time! Aircraft depend upon the atmosphere and its mass for their flight characterics--space doesn't have the atmosphere. Surface vessels move around on two-dimensional surfaces of water on a planet with a central gravitational field, spacecraft are in 3D space.... Terrestrial vehicles on on a planet with an atmosphere, and generally take oxygen from the atmosphere to use as oxidizer. Spacecraft are not in an atmosphere... the atmosphere also provides oxyge that people on surface vessels and in air vehicles and on land vehicles breathe.... and the atmosphere has weather, and the air can be corrosive (salt air), and there is drag.... spacecraft in space don't need streamlining, don't get corrosion from salt sea/salt air... ones in high orbit do have SCATHA issues....

I've heard some of what I considered extremely silly/misleading perceptions about space stuff, based on persons starting from excursion from surface naval fleet operations, or operations of air vehicle fleets, etc. And I pointed out, "space isn't a mission, it's a place/regime, and a very different environment from terrestrial/planetary surface, subsurface, or atmosphere. (I didn't tend to go further into orbital versus interplanetary etc., when considering primarily orbital versus terrestrial.)

Anyway, I do owe for the pointing out ab out the baggabe on the term "ship" [as oppose to "craft"] although, there are such things as
"airships" which do not have the same thought-baggage as "ship" thinking of "vessel which floats on water and has heritage of seamanship and enormous tail of associations." [Or, the newer usage of "ship" as abbreviation of "relationship" which e.g. steampunk fans and others use referring to their preference in character relationships, and the fans themselves as "shippers." It initially threw me the first times I saw those usages....]