Thought Experiment: A Realistic/Hard SF Galactic Empire

[Somes J]

No the reason why I consider 'Hard' science of the 'strictly what's possible today' overly restrictive is that aside from Warp Drives, you also cannot have:

Well that right there makes most* kinds of interstellar settings impossible, but who actually demands that? I mean futurism is a pretty huge theme in sci fi, and this kind of standard almost rules it out by default, since it demands literally no technological advancement whatsoever. I have never encountered anybody going "OMG not possible right now FAKE SoD broken!", at least not that I can remember.

* Although I think radio/laser communications between nearby stars could probably be done today if anyone was talking, so you could probably have a setting with modern or inferior tech and interstellar dialog, just not interstellar travel. There's probably at least a couple of potential interesting stories in that scenario.

Economy wise (also a 'sin' against physics, by the way), 'hard' science likes to have manned exploration beyond our solar system, besides the fact that there is nothing out there even remotely interesting to us until we invent replicators (if they were possible), because until we develop a post-scarcity economy, no one's going to fund or provide the ridonculous amount of resources required to make a manned flight out of the solar system possible.

Eh, I don't mind that so much because people are motivated by more than just doing the most economically rational thing. The Apollo program probably didn't make a whole lot of economic sense either and was more or less done just so we could say we did it before the Russians. The Great Pyramids probably represented an absolutely ridiculous amount of sweat and treasure for ancient Egyptian society and they were basically glorified headstones. I find it pretty easy to imagine humans doing stuff that would make very little sense to an economist-bot.

The assumption that an interstellar ship would necessarily be a huge resource commitment is also not really a given. It depends on a lot of economic and technological assumptions which in turn depend on a lot of factors, many of which are inherently unpredictable or simply do not have any one right answer period. There is no known inherent reason why, for instance, a deuterium-fueled fusion rocket would have to be outrageously expensive. Outrageously expensive is also very relative - ancient Romans would probably have been incredulous at the amount of metal needed to make a CVN, but it isn't outrageous to us. Similarly, a civilization commanding the resources of the solar system might well be able to contemplate things that would be outrageous to us, even if they aren't actually effectively postscarcity.

Incidentally your comment makes me feel better about the latest revision of my own pretentious ERECT SCI FI, where I eliminated the whole idea of interstellar colonies and just went with a Star Trek Federation-esque union of technologically advanced worlds. Although pretty much all these worlds are colonies of an ancient space empire that collapsed long ago, so I really just moved the problem back a little, but it's sort of supposed to be an exercise in writing cheesy space opera without magitech not 100% plausible this is the real future stuff so whatever.

Not a work of fiction, but the fans. Or even more so the non-fans and critics. The kind of people who poo-poo Star Wars/Star Trek/Babylon Five/Farscape for being fantasy (yes, yes it is, Science Fiction is a genre of fantasy), but laud 'hard' sci-fi for being 'realistic', even when it's more full of shit than the 'soft' stuff is.

Ah. From the "refuse to believe society can change" I'm guessing you may also be referring to the "the Federation is a communist dystopia!" crowd at SDN. Yeah, those people annoy the hell out of me to. Though actually I have somewhat the opposite annoyance with them. They won't bat an eyelid at 9999 gigaton lazors but any hint of society changing for the better and it's OMG FAKE UNREALISTIC SOD BROKEN! My suspicion is a lot of them are just uncomfortable with the suggestion that Spartafreedomerica might not be the bestest society feasible.

It may not be the most scientifically feasible solution, but at least it's a solution. Most 'Hard' science fiction stories either ignore the problem or aren't even aware that it exists.

Well, in fairness stuff like that might just be unmentioned background - if it's not important to the plot there's no particular reason to mention it, just like there's no particular reason to dwell extensively on the mechanics of cryosleep if it's not plot-relevant. It'd just be an unnecessary boring infodump that would interrupt the flow of the narrative and look like the writer wanting to show off how smart he is, which is pretty annoying.

[Somes J]

FYI: "Manned or unmanned flight to another planet in our solar system in less than a year."

This is laughably untrue; even the most cursory investigation would have realized that we've done it several times. Mars and Venus are both less than a year away. I'm pretty sure Mercury is too. Almost every, if not every last one of them, of probes sent to those planets arrived in less than a year. There's technology in the lab today (successful test flights, but not yet been used in production) able to make it to Mars in a just couple months.

I think some of the problems he suggests for interstellar spacecraft are a little exagerrated too. As I remember the Project Daedalus proposal dealt with impacting debris with a 50 ton metal shield, so they clearly didn't think it was a showstopper. Granted Daedalus had a top speed of ~.1 c, but you could still have an interstellar setting with that. A relatively reasonable-size radiator (dozens of meters on a side) could radiate away megawatts of waste heat, which should probably be plenty for survival systems*.

Not that there aren't plenty of difficulties in building a starship, realistically.

* Edit: come to think about it he might be talking about radiation, but space actually provides one possible neat solution for that: put it at the end of a long boom and let the inverse square law take care of most of it. There's no gravity except accelerational and little need for streamlining, so you could easily have a really long ship with the habitat at one end and the engines and reactor at the other.

[Blackwing]

Sorry, for manned or unmanned spaceflights to other planets in the solar system, I should've specified that I meant 'commuter flights'.

Yes you can get to Mars in under a year... When you launch your unmanned probe when the distance between Earth and where Mars will be when you arrive is optimal.
When Mars is on the other side of the Sun, however, trying to get there in a few months is still not possible. This does not stop many 'hard' sci-fi stories from having a monthly (or even weekly) commuter flight to 'the Mars colony' without anyone going 'Wait a second, this story is set on a date when Mars and Earth are on opposite ends of the sun, no matter what speed it goes, it'll have to slow down and turn towards Mars at some point and that means it would need to go at a fraction of c not attainable without magic'

At any rate, people are too quick to say that Star Trek uses magic (Star Wars, indeed, DOES use magic, but then again Star Wars isn't meant to be realistic), simply because it uses technology that's just as plausible as anything else, provided we can find a way to make it work.

Space Elevators, for instance, are not impossible according to our current tech because they're 'magic', but because we know for a fact that we can build one given enough resources and a material of sufficient tensile strength. We currently do not have such a material... But if we ever discover one, then we can build a Space Elevator.

The problem there is that we don't know whether we'll ever discover one and whether it is even possible for such a material to exist. Some parts of Physics tells us it should be possible, but other parts of Physics tell us that it may not be possible for us.

The same is true for a 'warp drive'. The physics say it CAN work, but we'd need to discover some tech that we have not yet discovered and that we may never discover, it's no less possible than Long-Term Cryonics or a Space Elevator.

In the end, for me, Science Fiction is about thinking up technologies that we may or may not ever possess and then thinking about how they would affect the world and the story. Whether those technologies are actually feasible in the next few generations is secondary to how well the handling of the technology within the story is done.

And sneering at 'soft' sci-fi for taking greater leaps is just as dumb as sneering at 'hard' sci-fi for not 'daring enough'.
The only time where 'sneering' of any kind is called for is when a soft sci-fi starts throwing around ridiculously powerful tech without anyone in the story blinking an eye or when hard sci-fi demands that not even the remotest conjecture is allowed, but still expects the story to take place in space.

[Destructionator]

Fun fact: I talked to an AI last night. My home cable went out - that's why my telecommute link broke - and I couldn't fix it myself. So I called the company.

This nice lady computer picks up the phone after a single ring and carries on just like a human representative would have. I read her my account info and explained the problem. It understood reasonably well - without me having to repeat myself or use normal multiple choice menus - and started taking steps to correct the problem.

The program on the other end was pretty flexible. I figure it listened for keywords in my problem explanation and then moved on with the script to a few simple fixes for that problem. Hilariously, the recording also included fake background sounds, like a keyboard being used when it said "please give me a moment to bring up your account". I laughed - it even got the right number of keypresses for my phone number, and understood when I said "niner" rather than "nine"! The pacing was a bit wrong though.

Anyway, it ended by sending a remote reset command to my cable box - definitely the wrong move, since my whole service was out, but maybe it didn't really understand the problem - and said if this doesn't fix it, call us back.


So I played along and called back when it didn't work. The computer says it remembers me and transfers me to a human in tech support. I actually forgot I was talking to a computer briefly and read from my own scripts of goodbyes ("thank you, you have a nice day too" kind of thing)



Anyway I'm transferred and put on hold. Finally someone picks up. This is where things get wonky.


While the computer was flexible and understanding, the human was rigid. He was clearly reading from a script and flow chart, with no latitude. One example:

"From left to right, tell me the status of each light on your modem. Blinking or solid."

"They are all off, except the far right one, which is blinking."

"OK, now from left to right, tell me the status of each light on your modem. Blinking or solid."

"Every light on the left is off. Only the one on the right is turned on. It is blinking."

"I need you to tell me from left to right if each light is blinking or solid."



Oh my god! The fucking lady computer understood this, but the human didn't? I started wondering if they transferred me to a less advanced AI instead.

After finally playing with the script (including all the things the computer already tried - and it had the decency to not do them again on the second call, but the human said "its standard procedure we have to do it every time"), he said it must be the amplifier (yup) so they'll send a guy out.


Between 8 and 5 the next day. So I might lose a full day of work.... but at least the guy got here early, and things went well. He swapped out the bad part in the locked box and problem solved.


But I just can't get over how the computer recording gave a more natural interface than the human representative! AI is already here. A full blown human intelligence isn't needed for a large number of tasks, and even when we have humans, we restrain them with scripts and procedures anyway, so it is scarcely any better than machines. Just slower, poorer listeners (didn't repeat anything to the computer, it knew. Human spelled my name wrong, even when I spelled it out phonetically ("Romeo, uniform...") Ugh. He must have been distracted.), but with the added benefit that they are a little better at the small talk. Small talk which increases the call time, and thus costs the company money.... but some people are more comfortable with it, so it isn't all bad.


Of course they'll still be those 1% of cases where some original thought is required and you want someone who can go off script, but that's what the AI already handles - it filters the easy calls out, leaving only the harder ones for people. The cable company could get a better AI, and better guidelines for its reps, but the idea is already in place. And the phone guys act as the second filter - solving things without needing to send someone on-site.

[Blackwing]

It's nice, but it's not AI.

When you have a problem appropriate to that programme's function, you're golden, present it with even the slightest variation, not so much..

For instance, let's say you're red-green colour blind and the programme asks you to identify whether a light is blinking red or green, your response is going to be 'I don't know'. At which point the programme will immediately tell you that it's going to transfer you to a human operator. This seems like a very smart thing to do for the programme ("Wow, it understands that it can't help me with this so it transfers me to a human being."), but instead it's that programme's pre-programmed response for 'Error, Error, unexpected response, does not compute'.

Likewise the number recognition wouldn't work in Dutch. The programme understands 'niner' because it expects only numbers and only listens to certain consonant+vowel combinations. 'woh' 'oo' 'ee' 'ou' 'fai' 'ih' 'eh' 'a' 'ain'.
If you'd told the programme 'I can't read this', the programme hears '(5 or 9) 7 3 6'.
In Dutch however, the 'a' sound that you have in eight, we have in 1, 2, 7 and 9. We have an 'ee' in 3 and 4. only 5, 6 and 8 are distinct in sound. This is why, when I call my bank and they need my number, I have to enter it into the keypad rather than tell a voice analyser, because there are still very few voice analysers accurate enough to tell Dutch numbers apart.

It's not smart, it's just very cleverly hidden stupidity. It's still very impressive, but much of what makes you go, 'oh it's very clever' is your own compliance with what the programme expect of you. It won't be true AI until, regardless of whether it's sentient, it no longer relies on pre-programmed responses.

[Destructionator]

It's nice, but it's not AI.
[...]
It's not smart, it's just very cleverly hidden stupidity. It's still very impressive, but much of what makes you go, 'oh it's very clever' is your own compliance with what the programme expect of you. It won't be true AI until, regardless of whether it's sentient, it no longer relies on pre-programmed responses.

By that definition, the human agent isn't intelligent either. Both man and machine do the job in almost the same way - both read from a script, handle pre-set responses, and if you go too far from it, both would probably break.

There's no practical difference between the two until you get into the 1% edge case. For computers, it is always the same. For humans, each one probably has a different area he can't handle, but everyone has one still.

At which point the programme will immediately tell you that it's going to transfer you to a human operator. This seems like a very smart thing to do for the programme ("Wow, it understands that it can't help me with this so it transfers me to a human being."), but instead it's that programme's pre-programmed response for 'Error, Error, unexpected response, does not compute'.

Likewise the number recognition wouldn't work in Dutch.

Humans do the exact same thing. Ever called someone in the tech support department then had a billing question? Sometimes the agent is helpful. A lot of times, it is "omfg you went off script, time to transfer". (The worst is when you have a tech support question, but don't word it the way they expect... and they transfer you just to get rid of you. At least the computer isn't a lazy asshole.)

The computer has a big advantage here though: it could be programmed to understand Dutch by uploading a translation pack, and then handle any number of callers. With a human, you'd have to hire someone else, or be transferred to the Dutch department. Who knows how long you'll be waiting for someone to become available.

It's not like the representative is likely to apply his intellect to understand the new language on the fly (if its even possible in the first place; he might just not have enough information, even if he were a god of smartness).

This is why, when I call my bank and they need my number, I have to enter it into the keypad rather than tell a voice analyser, because there are still very few voice analysers accurate enough to tell Dutch numbers apart.

Yea, but again, humans do this too. That's the whole reason things like phonetic alphabets exist. We have the same limits, just slid to somewhere else on the scale.




Anyway, some people insist on Strong AI being a complete drop-in replacement for a human character, though maybe with a few funny quirks. I don't really see a need for that - these limited scripts, when given adequate flexibility for its specific job, do just as well, with a much lower cost. Unless you specifically try to break it, the difference from a human drone is indistinguishable for the caller, so what does it matter if it is strong or weak ai?

You can apply the same logic to a lot of tasks, not just handling callers on a script. This is why I'm a lot looser with the definition of AI than a lot of sci fi fans. There's just no difference in most cases it actually handles.

[Destructionator]

it would need to go at a fraction of c not attainable without magic'

I hate it when anyone says "fraction of c", hard or soft. It almost certainly means the person hasn't really thought about it and just wants to sound cool.

A burn/flip/burn from Earth to Mars on opposite sides of the sun in one week requires delta-v of less than 0.5% the speed of light. (The average speed, of course, being half that or so.) Is that really a "fraction of c"? Technically, yes, of course, walking speed is technically a fraction of the speed of light. But is it meaningful to say that?

800 km/s is a far more usable number. Easier to write, easier to read, easier to relate to, easier to do math with. It wins hands down.

The same is true for a 'warp drive'. The physics say it CAN work

No, it doesn't. This is a common misconception of what the scientists actually did. I think I said this before, but they worked backward, not forward.

"I want this to work. How can we change the universe to make it happen?" instead of "This is the universe we have, how can we make this work?"


Such things are interesting to scientists because it can help them look at the theory in a new way, and can uncover something actually applicable to the real world, or maybe just get the author some fame among nerds who only read the title of their paper (if that), but it does NOT say "it is possible" and most certainly not that we can actually do it!



You seem to have two definitions of possible that you flip between. Sometimes it is: "what we actually do today and nothing else" and sometimes it is "what some obscure math might allow". Neither one is particularly useful.

I personally use a very conservative definition, but one that still allows for the future: if I had trillions of dollars at my personal command to put toward whatever I want in the next 20 years, what could we build? What projects that show promise in the lab could be finished? What battery of minor improvements might we make on all this stuff?

I usually don't get into things that haven't already gotten some seed study in the real world, but at the same time, I won't discard something just because it isn't already used in production. Sometimes, I'll even throw in pure magic; hard vs soft sci-fi isn't black and white.


Hence, it's my personal balance between science and fiction. If your's is different, that's cool, but no need to flame about it and certainly no need to say things that aren't true.

[Destructionator]

One last thing, on launch windows: what you call a limit, I call an opportunity. It's a classic race against the clock, in space.

Launch windows are one of the things that make alien invasions of Earth workable.


"we have two weeks left before it closes better work fast and take chances"

"can't launch yet. since we're stuck here for a while, up for some character development?"

"uhh we dont have 2 years worth of shit. wanna invade earth?"

[Somes J]

Yes you can get to Mars in under a year... When you launch your unmanned probe when the distance between Earth and where Mars will be when you arrive is optimal.
When Mars is on the other side of the Sun, however, trying to get there in a few months is still not possible.

Fun fact here: the kind of transfer orbit that's easiest in terms of energy (Hohmann*) is also the one that takes the most time. It's an elleptical orbit with the perigee at the lower orbit and the apogee at the higher orbit, so the ship needs to complete a full half an orbit to make the transfer, and it crosses the maximum distance between the two bodies (2.5 AU in the case of Earth and Mars). For a shorter transfer you need a more elleptical orbit with apogee or perigee below or above the destination orbit, which takes more energy.

If you're getting to the destination in less than the maximum possible transit time it means you're actually using a harder transfer orbit.

Reference.

* This is not exactly true as there's a trick called the Interplanetary Transport Network that can do lower energy transfers but it's ridiculously slow: two years to get from the Earth to the Moon.

This does not stop many 'hard' sci-fi stories from having a monthly (or even weekly) commuter flight to 'the Mars colony' without anyone going 'Wait a second, this story is set on a date when Mars and Earth are on opposite ends of the sun, no matter what speed it goes, it'll have to slow down and turn towards Mars at some point and that means it would need to go at a fraction of c not attainable without magic'

Actually a transfer from the farthest points Earth and Mars get from each other is the easiest one, as I said above. It's the transfers when they're close together that will be hard because you need to put the spacecraft on a highly elleptical orbit with a apogee way above Mars orbit to hit Mars (or perigee way below Earth orbit, if you're launching from Mars).

Also, the solar system escape velocity from Earth is 42.1 km/s, which as Destructionator pointed out is a "fraction of c" only in the sense that it is a little more than .0001 c. At this point you can more or less give the finger to orbits as your ship is no longer even bound to the sun. In practice I doubt you'd even need that much to be able to do nearly year-round transfers; you just need to be able to put the ship in a very highly elleptical comet-like orbit, after all. Some electric rockets like VASIMR might comfortably be able to attain dozens of km/s delta V. You might say they don't exist today and are therefore magic but they're far less magic than something like warp drive: people are seriously contemplating building some of them today. Granted such rockets tend to be so low-thrust that might actually be a limitation for some very close launch periods, but not having Earth-Mars transfers restricted to Hohmanns and near-Hohmanns is not really all that ridiculous.

I'd also like to point out that several space probes actually are on escape trajectories from the solar system in real life, although IIRC they got gravitational boosts off the giant planets to do it.

Edit: come to think of it maybe what you complain about is the actual Mars trip being too short... do you have any stories in mind where this problem shows up, so I can know how long it took?

Space Elevators, for instance, are not impossible according to our current tech because they're 'magic', but because we know for a fact that we can build one given enough resources and a material of sufficient tensile strength. We currently do not have such a material... But if we ever discover one, then we can build a Space Elevator.

Eh, I think space elevators (from Earth anyway) are massively overrated anyway. Even if we could build one I think mass drivers make a lot more sense as a launch system. I think they're one of those things that get talked about a lot more because they're well known than because they're actually such a great idea.

The same is true for a 'warp drive'. The physics say it CAN work, but we'd need to discover some tech that we have not yet discovered and that we may never discover, it's no less possible than Long-Term Cryonics or a Space Elevator.

This is where I fundamentally disagree with you. You seem to have this standard, which I find bizarre, that either you ban everything that can't be done right now or everything is equally plausible so anything goes. There are degrees of magicalness (which I think is a terrible term here but I'll use anyway). A manned VASIMR rocket might be a little magical but not much. A fusion rocket is more magical. A fusion rocket capable of reaching .1 c is more magical yet. A Bussard ramjet is still more magical. And warp drive is more magical than that.

Incidentally as I've said before from what I remember the way warp drive works is such that even if the physics actually allows it (and as Destructionator points out this is actually not a given) you'd require a lot more magical dancing panda technology before you could actually build one because of the literally astronomical energy requirements.

[Blackwing]

What the human with the script does and what he's capable of are two different things, so no a pre-programmed computer is not equal to a human being in this respect. A computer is much easier to 'break' in that respect. Tell the computer 'I think you're being helpful' and it will shut down, tell the person the same thing and while it may not be part of the script it will not make them unable to respond.

All of that doesn't matter any way because that's just in a very limited context. The human will also transfer you if you go off script IF they're simply call centre workers who don't know anything about the tech in question, that doesn't mean that the computer is somehow an AI, just because a human being runs through a script.

When you give a human call centre worker a different job in a different call centre with a different script, they're still able to function. The telephone operator programme can't without being completely rewritten.

At any rate the loose definition of an AI, and at any rate the condition that a system must meet to be called an AI, is that it must be able to solve problems it was not programmed with a solution for and it must do so on it's own.

Essentially something is not an AI when, like the telephone operator programme, it is a simple mesh of 'IF(preprogrammed input)THEN(preprogrammed output);ELSE(preprogrammed output);' switches.

In order to qualify as an AI it needs to be able to solve a problem that it was not specifically programmed to handle in a way it was not specifically programmed to handle it. It needs to be able to learn from past failures and successes in order to provide a solution more efficiently, successfully and/or quickly (ultimately all three at the same time in comparison to it's earlier attempts) and it needs to be able to apply some form of logic to do this (simply randomly attempting possibilities and discarding ones that don't work is a good start to building or teaching an AI, but it only becomes an actual AI once it's able to analyse why the things that work work and why the things that don't work don't).

It doesn't need to be sentient to do this, but it does need to do this to be an AI.

That what makes it an AI. That's not the Sci-fi opinion, that's the ACTUAL requirement.

So yes, useful programme, very clever. Not an AI.

A burn/flip/burn from Earth to Mars on opposite sides of the sun in one week requires delta-v of less than 0.5% the speed of light. (The average speed, of course, being half that or so.) Is that really a "fraction of c"? Technically, yes, of course, walking speed is technically a fraction of the speed of light. But is it meaningful to say that?

800 km/s is a far more usable number. Easier to write, easier to read, easier to relate to, easier to do math with. It wins hands down.

I may be doing the math wrong here (and I mean that, math is not my strong point), but is that based on the absolute distance between the planets? I get an average velocity of 600km/s to make it to Mars in a straight line (through the sun, not an option). With a bit of guess work, my assumption is that your average velocity is based on a curve around (not closely around, of course). Is that the case?

At any rate, that's still a whole order of magnitude greater than the fastest man-made object in space so far.

Lastly, this made me realize something:

One last thing, on launch windows: what you call a limit,

No, no, no, I think I get the misunderstanding now...

That list of things I gave was not a list of things that I consider limitations on what a 'hard' sci-fi story should stick to and thus why I think that 'hard' sci-fi is limited.
That list is a list of things that are either physically impossible with any tech we currently have or can prove will work.
Space Elevators, as I mentioned, are physically possible, but only if we find a way to manufacture hypothetical material with certain properties of which we cannot currently devise a feasible way to make them work. Carbon nano-tubes have been suggested as a possibility, but this depends on whether the properties we've observed on a microscopic level will be applicable on a macroscopic level.

And the reason why I gave that list is that these are all things that I've seen in science fiction short stories on the net that were lauded as being good, realistic ('hard') science fiction because 'everything in the story is completely realistic according to our modern understanding'.
Mostly on FF.net's original fiction brother (Fictionpress.com) and DeviantArt's lit section. Yes I'm aware those aren't exactly the most highbrow places in any case. But still, a good deal of snobbery around.

Personally I rather enjoy 'modern tech and what's possible' Sci-fi set on Earth, I just get a bit miffed when unthinking authors think that Space Elevators and Cryonic Suspension fit that description and get rewarded for it with praise affirming that notion by the same sneering assmonkeys who two days earlier berated another piece for not being realistic enough for having an FTL drive in it.

Also note that I said that a 'Warp Drive' is possible. Not the Warp Drive (the star trek one). Though I guess that's my mistake for not making that clearer. (On the other hand Somes J used 'Lighthugger' as a general term too, rather than referring to the Revelation Space ones from which the term comes. For reference, Revelation Space 'Lighthuggers' have technology that allow them to IGNORE THE SECOND LAW OF THERMODYNAMICS!)

[Somes J]

I hate it when anyone says "fraction of c", hard or soft. It almost certainly means the person hasn't really thought about it and just wants to sound cool.

What I hate is when people throw around high fractions of c as if they're nothing special, clearly probably having no clue what it actually means.

Same with terawatt, or gigaton, or other similar huge numbers.

My favorite is when you'll get guys in places like SB and SDN who poo-poo ridiculous power as pathetic. I remember some guy on SDN once saying something about the Enterprise's reactor being unable to generate a "measly terawatt" as if that was somehow pathetic and I was all lol. Your space-limited spacecraft-carried reactor can't generate like 1000X the power of the biggest real life nuclear power plants OMG U SO WEAK!

I think a lot of sci fi nerds have very little real life perspective on science from which to perform sanity checks on their fantasies.

[Somes J]

Also note that I said that a 'Warp Drive' is possible. Not the Warp Drive (the star trek one).

I thought you were talking about the real life Alcubierre/Van Den Broeck warp drive proposals.

On the other hand Somes J used 'Lighthugger' as a general term too, rather than referring to the Revelation Space ones from which the term comes. For reference, Revelation Space 'Lighthuggers' have technology that allow them to IGNORE THE SECOND LAW OF THERMODYNAMICS!

I like the term because it's less clunky than "high c fractional spacecraft".

Incidentally if you're talking about the cryoarithmatic engines I don't think they're an essential component to Reynolds-verse lighthuggers. They might have other ridiculous stuff though - the books I've read didn't go into much detail on their operation. And even at the most realistic interpretation they seem to be some kind of Bussard ramjet which is about the most "magical" drive system you can have without going into stuff that outright breaks physics as we know it.

[Destructionator]

Tell the computer 'I think you're being helpful' and it will shut down, tell the person the same thing and while it may not be part of the script it will not make them unable to respond.

Unless you get the guy I talked to last night!

At any rate the loose definition of an AI, and at any rate the condition that a system must meet to be called an AI, is that it must be able to solve problems it was not programmed with a solution for and it must do so on it's own.

The dictionary also tends to include things like expert systems, voice recognition, natural language translation, image processing, and quite a bit more under the umbrella of AI.

And like the phone story is supposed to be saying: put those things together, and you have a machine able to do a human's job quite well. The difference between actually being smart and being really good at pretending to be smart just doesn't matter most the time in practice.


Limited AI like this can be a plus for the story too: it is there to help, but not to star. The real spotlight is still on your human characters.

I may be doing the math wrong here (and I mean that, math is not my strong point), but is that based on the absolute distance between the planets? I get an average velocity of 600km/s to make it to Mars in a straight line (through the sun, not an option). With a bit of guess work, my assumption is that your average velocity is based on a curve around (not closely around, of course). Is that the case?

Basically yeah. I added in some extra to allow for slowing down and changing direction too. (If you draw a picture, Earth is going up and Mars is going down, so you've gotta cancel the planet's motion too with your rocket.)

I also did a lot of intermediate rounding to make the math simpler for my head, but we're having an internet discussion not engineering a real mission so it's good enough for me that way. Probably plus or minus 50% is reality.

At any rate, that's still a whole order of magnitude greater than the fastest man-made object in space so far.

Yea, it's insanely fast. I prefer to avoid such things in general.

And the reason why I gave that list is that these are all things that I've seen in science fiction short stories on the net that were lauded as being good, realistic ('hard') science fiction because 'everything in the story is completely realistic according to our modern understanding'.

Well if you want to say some specific examples of hard sf suck, I'll surely agree if I read it! My conservative viewpoint is rare among Internet authors. In other threads, I've been the one to attack things like space elevators for being silly.

Your other posts just went a bit too far.

Also note that I said that a 'Warp Drive' is possible. Not the Warp Drive (the star trek one). Though I guess that's my mistake for not making that clearer.

No, I know what you're talking about, and that's what I'm attacking. An actual Star Trek warp drive doesn't bother me; it knows its place is to enable plots and doesn't pretend to be real science. (Maybe some hard core fans do, but mostly it isn't talked about that way.)

The general relativity based warp drives are what I'm discussing - they are real science, but worked backward. I'll agree that the kind of hard sf wanker who says "omfg a paper I didn't understand agrees with me says its possible so SUCK IT BITCHES" are losers.

[Destructionator]

Actually a transfer from the farthest points Earth and Mars get from each other is the easiest one, as I said above.

Not quite. This depends on the specific transfer, but for Earth/Mars, the Hohmann launch window is actually when they are about 45 degrees apart.

You might say they don't exist today and are therefore magic but they're far less magic than something like warp drive: people are seriously contemplating building some of them today.

VASMIR is actually better than that - it's been test fired in the lab, with several prototypes built. It is slated to be test-fired on the ISS in 2014. Hopefully the station will still be there!

http://www.aviationweek.com/aw/generic/ ... %20Strides

Last fall, the VX-200 Vasimr prototype reached full-power plasma thrust under the control of a superconducting magnet in vacuum conditions. The achievement marked a critical milestone in Chang-Diaz’s long-running efforts to develop an electric propulsion drive that could one day transport humans to Mars in 39 days.

They've made an even bigger prototype just since the time I first read about it (linked from Atomic Rockets). Very exciting.


Ion drives have already been used in space too.

[Blackwing]

Fun fact here: the kind of transfer orbit that's easiest in terms of energy (Hohmann*) is also the one that takes the most time. It's an elleptical orbit with the perigee at the lower orbit and the apogee at the higher orbit, so the ship needs to complete a full half an orbit to make the transfer, and it crosses the maximum distance between the two bodies (2.5 AU in the case of Earth and Mars). For a shorter transfer you need a more elleptical orbit with apogee or perigee below or above the destination orbit, which takes more energy.

If you're getting to the destination in less than the maximum possible transit time it means you're actually using a harder transfer orbit.

I don't understand what you're talking about...
I hate to be so blunt about it, but this has fuck all to do with the matter. It's a fun fact, yes, but it supports what I said when it comes to Earth-Mars commuter flights. A Hohmann Orbit from Earth to Mars is only possible once every ~2 years and takes over 8 months to complete. Half the elliptical orbits required to get to Mars from Earth efficiently run through the sun (or near enough to be infeasible). And when Mars is opposite the Earth from the Sun, the only way to get there without orbiting the sun completely is to go in the opposite direction to the Earth's orbit around the Sun, at which point you need infeasible amounts of thrust to even STOP your craft, let alone getting it moving in the right direction.

And when I said 'weekly' I did not mean 'one leaves every week', I mean 'each one takes a week to get to Mars'.

Actually a transfer from the farthest points Earth and Mars get from each other is the easiest one, as I said above. It's the transfers when they're close together that will be hard because you need to put the spacecraft on a highly elleptical orbit with a apogee way above Mars orbit to hit Mars (or perigee way below Earth orbit, if you're launching from Mars).

Easiest only in the sense that you don't need to accelerate very much. It still takes over a year to get there. So that doesn't comply with the 'get to Mars in a week' thing.

Also, the solar system escape velocity from Earth is 42.1 km/s, which as Destructionator pointed out is a "fraction of c" only in the sense that it is a little more than .0001 c. At this point you can more or less give the finger to orbits as your ship is no longer even bound to the sun. In practice I doubt you'd even need that much to be able to do nearly year-round transfers; you just need to be able to put the ship in a very highly elleptical comet-like orbit, after all. Some electric rockets like VASIMR might comfortably be able to attain dozens of km/s delta V. You might say they don't exist today and are therefore magic but they're far less magic than something like warp drive: people are seriously contemplating building some of them today. Granted such rockets tend to be so low-thrust that might actually be a limitation for some very close launch periods, but not having Earth-Mars transfers restricted to Hohmanns and near-Hohmanns is not really all that ridiculous.

I'd also like to point out that several space probes actually are on escape trajectories from the solar system in real life, although IIRC they got gravitational boosts off the giant planets to do it.

Yes. Specifically the Helios 2 probe, the fastest object man has ever made ever, reached nearly twice that... Over the course of months... By slingshotting the sun from ~0.3 AU, too close for humans to survive. Not to mention it accelerated at enough g's to turn humans into jelly (not literally) for part of that.

Edit: come to think of it maybe what you complain about is the actual Mars trip being too short... do you have any stories in mind where this problem shows up, so I can know how long it took?

Off hand mentions in a number of them of trips to Mars in a month, with one or two outliers having it within a week. Can't recall the specific names, but I'll look around.

Eh, I think space elevators (from Earth anyway) are massively overrated anyway. Even if we could build one I think mass drivers make a lot more sense as a launch system. I think they're one of those things that get talked about a lot more because they're well known than because they're actually such a great idea.

Mass Drivers though, also require massive amounts of energy to operate. A Space Elevator, if we can make one, would require minimal energy input compared to a Mass Driver (especially if we use it to transport materials back to earth with a greater weight than what we sent up there). A Space Elevator also allows whatever material it carries up to be carried down again without a special landing mechanism, something a Mass Driver can't.
Which is the real reason why they are so popular. A space elevator is more than a really long tube.

This is where I fundamentally disagree with you. You seem to have this standard, which I find bizarre, that either you ban everything that can't be done right now or everything is equally plausible so anything goes. There are degrees of magicalness (which I think is a terrible term here but I'll use anyway). A manned VASIMR rocket might be a little magical but not much. A fusion rocket is more magical. A fusion rocket capable of reaching .1 c is more magical yet. A Bussard ramjet is still more magical. And warp drive is more magical than that.

Incidentally as I've said before from what I remember the way warp drive works is such that even if the physics actually allows it (and as Destructionator points out this is actually not a given) you'd require a lot more magical dancing panda technology before you could actually build one because of the literally astronomical energy requirements.

Actually a manned VASIMR is much much magic. Due to heat. Because once again the biology of space takes a backseat to the physics of space. 'This goes fast, if we put people on it, those people will go fast'. Except that a VASIMR is not suited to manned spaceflight. Not that I'm saying you're intentionally ignoring the biology aspect of the thing.

[Destructionator]

A Hohmann Orbit from Earth to Mars is only possible once every ~2 years and takes over 8 months to complete.
[...]
Easiest only in the sense that you don't need to accelerate very much. It still takes over a year to get there.

Which is it, 8 months or over a year?

Half the elliptical orbits required to get to Mars from Earth efficiently run through the sun (or near enough to be infeasible).

Not true.

It still takes over a year to get there.

Not true.

Not to mention it accelerated at enough g's to turn humans into jelly (not literally) for part of that.

Not true.

Mass Drivers though, also require massive amounts of energy to operate.

Not true.

A space elevator is more than a really long tube.

Not true; though a little nitpicky: it isn't a tube at all.

Actually a manned VASIMR is much much magic.

Untrue.



If you'd just stick to saying things that are true, you'd get a lot less resistance.

[Blackwing]

Which is it, 8 months or over a year?

I meant if you extrapolated a Hohmann-style elliptical orbit to get to Mars when it's on the opposite end of the Sun from Earth, that takes over a year (and by the time you get there, Mars has already been at an actual Hohman appropriate angle and past that.

Not true. 1

'Efficiently' in the sense of the exact Hohmann Orbit's curve (rather than a similar orbit adjusted for safety). That is to say plotting the Hohmann Orbit's curve exactly when Mars is not at a ~45 degree angle to Earth. Half may be exaggerated, but a good deal of those run close enough to the sun to kill the crew or through the sun completely. Which is why the Hohman Orbit only works at that specific ~45 degree angle.
So yes, true.

Not true. 2

It does take more than a year to get to Mars if you launch when the Earth is opposite to Mars across the Sun.
So true.

Not true. 3

Nice rebuttal? If I calculated this correctly, it pulled about 8 g's in centrifugal motion which would over the course have put transverse, positive longitudinal and negative longitudinal stress on humans. The latter of which would cause brain haemorrhaging even though the earlier two are largely harmless.

Not true. 4

Yeah they do. Unlike a maglev train, a Mass Driver propelled spaceship can't use regenerative breaking. That means it takes a massive amount of energy (especially compared to a Space Elevator) to get something up. We're talking about a Earth-to-orbit Mass Driver suitable for manned craft here, not a space-based one for cargo.

Not true. 5 (nitpicky)

It can be if you want to vacuum pump it to reduce drag and conserve energy. (In fact, due to the weight-to-strength ratio a hollow structure is more feasible for a space elevator any way.... Plus if we make it out of carbon nanotubes it will be nothing BUT tubes )

Untrue. (for variety?)

Yeah it is. Heat management is the VASIMR's greatest problem. Adding a crew requiring a controlled temperature (as opposed to an unmanned vessel, which can simply turn the engine and the powerplant off to give it time to radiate away heat) does not help. It's not so much that it's IMPOSSIBLE to have a manned craft using a VASIMR engine, but it's better suited to keeping a large structure like the ISS in orbit through periodic, short duration bursts than powering a manned craft with more lengthy operation. That is if it doesn't fry every piece of equipment on the ISS when they turn those electro-magnets on.

If you'd just stick to saying things that are true, you'd get a lot less resistance.

You can't penetrate my meditative mantra... Ohmmmmmm... Ohmmmmmm... (I am such a geek)

[Destructionator]

I meant if you extrapolated a Hohmann-style elliptical orbit to get to Mars when it's on the opposite end of the Sun from Earth, that takes over a year (and by the time you get there, Mars has already been at an actual Hohman appropriate angle and past that.

So a trip is sub-optimal if your mission planner doesn't know what he is doing? Come on.

That is to say plotting the Hohmann Orbit's curve exactly when Mars is not at a ~45 degree angle to Earth. Half may be exaggerated, but a good deal of those run close enough to the sun to kill the crew or through the sun completely.

You have no comprehension of how big space is, do you? You apparently don't even know what a Hohmann orbit is. TIP: a Hohmann orbit never goes lower than the lowest of its destinations (so in this case, Earth).

If Mars isn't in the right place, you don't hit the sun. You just need to speed up or slow down to catch up to it. Hence it takes more delta-v (and hence propellant) to make up for the difference.

Which is why the Hohman Orbit only works at that specific ~45 degree angle.

No, it does so Mars is waiting right there when you arrive. Anything else would require additional fuel to make up for the difference, and with costs as high as they are today already, that's undesirable.

Nice rebuttal? If I calculated this correctly, it pulled about 8 g's in centrifugal motion which would over the course have put transverse, positive longitudinal and negative longitudinal stress on humans. The latter of which would cause brain haemorrhaging even though the earlier two are largely harmless.

Show your work. Where did this happen?

(answer: never, and even if it did, it wouldn't affect the crew unless it came directly from the engine... which an ion engine cannot possibly do.)

That means it takes a massive amount of energy (especially compared to a Space Elevator) to get something up.

Nope, it's exactly the same. There ain't no such thing as a free lunch. The best the space elevator can do is get paid back (mostly) when you return.

But, the amount of energy is quite small anyway, so it doesn't make economic sense to worry about it. ~60 MJ/kg, or 8 kWh / lb (rounding up also takes care of electromagnet inefficiencies). At the rates I pay for commercial electricity, that's about $1 / pound.

It can be if you want to vacuum pump it to reduce drag and conserve energy. (In fact, due to the weight-to-strength ratio a hollow structure is more feasible for a space elevator any way.... Plus if we make it out of carbon nanotubes it will be nothing BUT tubes )

Hilarious. About 0.1% of the journey is even done with significant air around it.

Heat management is the VASIMR's greatest problem.

You don't think the rocket engineers, physicists and astronauts who designed and built the thing don't know that? Seeing how it's been fired at full power, don't you think they know it much better than you do?

Adding a crew requiring a controlled temperature (as opposed to an unmanned vessel, which can simply turn the engine and the powerplant off to give it time to radiate away heat) does not help.

No, it actually makes no difference at all. The crew component is separate from the engine anyway. It's not like the heat will magically fly over there.

By the way, the 200 kW engine has about a 65% efficiency and comes in pairs. That's a total of about 130 kW of waste heat. About three square meters of metal radiator area will get rid of that as it is produced. It's a pretty easy problem.

It's not so much that it's IMPOSSIBLE to have a manned craft using a VASIMR engine, but it's better suited to keeping a large structure like the ISS in orbit through periodic, short duration bursts

These things are made for long burns.

That is if it doesn't fry every piece of equipment on the ISS when they turn those electro-magnets on.

Again, you don't think the people involved in the project know about these things? (btw I doubt it's even a problem at all in reality, but even if it is, it is apparently solved - would these armies of professionals have approved the project if they knew it was pissing billions worth of equipment down the drain?)

[Somes J]

And when Mars is opposite the Earth from the Sun, the only way to get there without orbiting the sun completely is to go in the opposite direction to the Earth's orbit around the Sun, at which point you need infeasible amounts of thrust to even STOP your craft, let alone getting it moving in the right direction.

Earth's orbital velocity is ~30 km/s; it's a fairly significant velocity and pretty huge by chemical rocket standards but nothing inconcievable. VASIMR could do it easily, for instance.

And when I said 'weekly' I did not mean 'one leaves every week', I mean 'each one takes a week to get to Mars'.

Ah, I was somewhat confused about what you were complaining about.

OK, that's a pretty high velocity (when they're at opposite ends of the sun ~600 km/s) but by the standards of most sci fi it's still fairly reasonable. I could easily buy an advanced fusion drive being able to do that, or some sort of external laser propulsion system (the latter is probably "harder"). It requires technology way above anything we can build, which has a chance of turning out to be impossible, but there's nothing explicity physics-breaking. I think you could have such drives and still legitimately call it hard SF. Not as hard as if more conservative drive systems were being used, perhaps, but still fairly hard.

Heck, any hard SF setting that wants interstellar travel is probably going to need much faster ships.

As far as them always taking the same time to reach Mars ... yeah, that would be just authorial laziness or ignorance. I'm not going to defend that.

Off hand mentions in a number of them of trips to Mars in a month, with one or two outliers having it within a week. Can't recall the specific names, but I'll look around.

Mars in a month at its closest approach to Earth I get ~30 km/s, just assuming a straight line path. For Mars in a week at closest approach I get 124 km/s. That's really not that unreasonable, it's within reach of (say) VASIMR.

Mass Drivers though, also require massive amounts of energy to operate. A Space Elevator, if we can make one, would require minimal energy input compared to a Mass Driver (especially if we use it to transport materials back to earth with a greater weight than what we sent up there).

It's actually a bit of a myth that getting into space period is expensive because it takes enormous amounts of energy. The 32 MJ/kg figure gets quoted a lot as if it's some obscene amount of energy but actually going by that with the energy of a large nuclear power plant you could put 4000 tons into orbit a day.

Meanwhile the space elevator is a much bigger engineering challenge and will probably have a much smaller handling capacity.

Actually a manned VASIMR is much much magic. Due to heat. Because once again the biology of space takes a backseat to the physics of space. 'This goes fast, if we put people on it, those people will go fast'. Except that a VASIMR is not suited to manned spaceflight.

Atomic Rockets gives the energy requirement for VASIMR as 10 megawatts. That magnitude of waste heat might require a relatively large radiator (maybe in the 100 X 100 m range assuming a temperature of ~500 K) but hardly magic.

There is also the option of putting the habitat at the end of a long boom. The inverse square law is your friend.

[Destructionator]

A space elevator IIRC can lift about 30 tons a day on average (with each load taking a week to get up!)... absolutely pathetic compared to a mass driver. Or rockets, for that matter.

The only good thing with a space elevator is that you can cut them in the opening moves of WW3. Like killing a redshirt to show shit has gotten serious.

edit: and i guess the view is kinda cool. It'd be a leisurely ascent.

[Somes J]

Not quite. This depends on the specific transfer, but for Earth/Mars, the Hohmann launch window is actually when they are about 45 degrees apart.

Yeah I was thinking of saying something like "mind you it's actually more complicated because you're aiming for where the planet will be not where it is now" but I figured I didn't want to add any more extraneous information.

If you look at the picture here though, you can see what I mean.

[Destructionator]

Ah yes, indeed.

[Blackwing]

No, it actually makes no difference at all. The crew component is separate from the engine anyway. It's not like the heat will magically fly over there.

By the way, the 200 kW engine has about a 65% efficiency and comes in pairs. That's a total of about 130 kW of waste heat. About three square meters of metal radiator area will get rid of that as it is produced. It's a pretty easy problem.

Plus the heat produced by the crew, plus the heat produced by the power plant supplying power to lifesupport, plus the heat produced (ironically) by the system keeping the crew at comfortable temperature. Plus the heat radiated in from nearby planets.

The problem that I have with people taking a "we'll just plop some radiators on there" approach is that the larger your radiator, the MORE heat you will take in when in direct view of the sun (i.e. anywhere not in the shadow of a planet) and the more heat you will take in as it is radiated away from a planet.

A space elevator IIRC can lift about 30 tons a day on average (with each load taking a week to get up!)... absolutely pathetic

Yeah, because that's so much more pathetic than using a 2000 ton rocket to launch a 100 ton space shuttle.
Besides, those are the specifications for what we'd be able to do in 2014.

[Destructionator]

Plus the heat produced by the crew, plus the heat produced by the power plant supplying power to lifesupport, plus the heat produced (ironically) by the system keeping the crew at comfortable temperature.

All negligible.

The crew is about 100 W a pop. Even with 100 people (huge huge), 10 kW is a small fraction of the engine's waste.

The power plant's heat is radiated from itself; the most likely one is solar panels which get a hot surface, but don't transmit that heat to the rest of the ship.

And keeping the cabin at a temperature costs virtually nothing extra; maybe just running a steam pipe.

Plus the heat radiated in from nearby planets.

Only a problem for the early and late parts of the trip, and considerably less than what they are already dealing with.

The problem that I have with people taking a "we'll just plop some radiators on there" approach is that the larger your radiator, the MORE heat you will take in when in direct view of the sun (i.e. anywhere not in the shadow of a planet) and the more heat you will take in as it is radiated away from a planet.

Don't point your radiators at the sun!

Yeah, because that's so much more pathetic than using a 2000 ton rocket to launch a 100 ton space shuttle.

A space elevator is hundreds of thousands of tons, if not more, that already need to be in space!

It'd have to run continuously for several decades just to pay for itself. Most proponents handwave it by saying it will be spun by robots in space and hence avoid this cost... too magical for my blood.

[Somes J]

Yeah, because that's so much more pathetic than using a 2000 ton rocket to launch a 100 ton space shuttle.

But we're not talking about space elevator vs space shuttle, we're talking about space elevator vs mass driver. The mass driver isn't limited to lifting a very small fraction of its mass over a period of days. It can probably easily beat the space elevator very handily in terms of handling capacity for this reason. It's also much easier to build since it's constructed entirely on the ground, and is comparatively tiny (even the hundreds of km long mass drivers required for an acceleration gentle enough for manned capsules are a small fraction of the length of the space elevator cable - and don't have to be entirely constructed from incredible superstrong materials).

They're not the end of potential competition for space elevators either. And space elevators have a lot of problems - maintenance will be a headache, every sattelite except those in weird or very high orbits will have to actively manuever around it - yes they're solvable but they're issues that other systems don't have to deal with (though mass drivers have their own problems). I just generally think the space elevator gets talked about out of proportion to its actual promise and usefulness because a lot of people are familiar with it and unfamiliar with the alternatives.

Besides, those are the specifications for what we'd be able to do in 2014.

I'm pretty skeptical of that - IIRC the space elevator requires thousands of tons of carbon nanotubes of a quality that we still struggle to make tiny quantities of in laboratories.