No Faster-Than-Light Travel

[DSMatticus]

Trouble is, 'near c' is still basically useless for any particular human being wanting to travel the galaxy, even if it's not a problem for machine that don't, like need to eat or have a relatively brief lifespan.

This is what I figure will happen. The human brain is data; eventually we'll figure out the mathematical model that represents it, and we will put that in a computer and fire it into space STL running on an efficient fuel source, possibly solar or nuclear. Or we'll come up with completely artificial, non-human-like intelligence, and fire that into space instead.

Either way, I'm fairly confident that in a century or two or so we'll be able to create an intelligence that can travel the stars at sub-light speeds and spread its form of civilization. Whether or not we'll choose to, I have no idea. There's certainly no incentive. But from an engineering perspective, it's a lot more feasible than a generation ship, because of the survival/logistic requirements: electricity instead of maintaining a human-habitable environment. It's also much easier on the "where should we land?" bit, because again: much wider range of environmental tolerance and much less strenuous survival requirements.

[fectin]

Does nuclear even work in space? I thought you had to have some pretty amazing cooling to not die.

[Kaelik]

Does nuclear even work in space? I thought you had to have some pretty amazing cooling to not die.

The temperature outside your ship being exactly zero Kalvin provides some sexy cooling.

[angelfromanotherpin]

The temperature outside your ship being exactly zero Kalvin provides some sexy cooling.

You fail physics forever.

[DSMatticus]

Does nuclear even work in space? I thought you had to have some pretty amazing cooling to not die.

The idea's been tossed around. I think we've built viable models of unmanned spacecraft using it, but nothing concrete. Small reactors, big radiator panels.

The temperature outside your ship being exactly zero Kalvin provides some sexy cooling.

To elaborate on angelfromanotherpin, the mechanism of heat loss you're referring to is conduction/convection. It only works when you actually contact particles, but space is a near-vacuum. If you don't touch particles, you can't give them your heat. Wind chill is a nice example of the opposite effect. The only way to lose heat in space is to radiate it off as a form of energy, and that happens at a rate based on surface area and material and some other stuff. I.e., you stick giant radiator panels on your ship, and direct your heat to them.

Edit: or maybe he meant the exactly zero kelvin thing. That would make sense too.

[Kaelik]

To elaborate on angelfromanotherpin, the mechanism of heat loss you're referring to is conduction/convection. It only works when you actually contact particles, but space is a near-vacuum. If you don't touch particles, you can't give them your heat. Wind chill is a nice example of the opposite effect. The only way to lose heat in space is to radiate it off as a form of energy, and that happens at a rate based on surface area and material and some other stuff. I.e., you stick giant radiator panels on your ship, and direct your heat to them.

Edit: or maybe he meant the exactly zero kelvin thing. That would make sense too.

I am aware that depending on how you draw the area, and chance, it might be either zero or something absurdly high.

But correct me if I'm wrong, the outside temperature has an effect on radiation as well.

[DSMatticus]

I am aware that depending on how you draw the area, and chance, it might be either zero or something absurdly high.

Well, an area can't have a temperature. That's an undefined operation, like asking the volume of a circle. Temperature is actually a description we apply to particles, whether individually or as a group. If an area has no particles, its temperature is undefined. You can find particles in space that have very high temperatures, and particles that have very low temperatures, and when we say "space's temperature is blah," what we mean is that the average temperature of the particle you encounter is blah.

But correct me if I'm wrong, the outside temperature has an effect on radiation as well.

Nope. Three factors: the radiating body's temperature, the radiating body's surface area, and the tendency of the radiating body to radiate.

[RadiantPhoenix]

The heat flux from radiation is εσA * Δ(T^4), where ε is a number based on what your radiator is made of, σ is a constant, A is the surface area, and T^4 is the absolute temperature raised to the 4th power.

IIRC, the actual temperature that represents space is about 3 K, not 0 K, but that's close enough to zero as makes no difference for estimation.

[angelfromanotherpin]

Even intergalactic space has a temperature of between 2 and 3 Kelvin. That's the cosmic microwave background radiation.

I don't believe that external temperature is a factor in radiative heat dissipation rates. I'm pretty sure that's based on absolute and not relative temperature.

[RadiantPhoenix]

I don't believe that external temperature is a factor in radiative heat dissipation rates. I'm pretty sure that's based on absolute and not relative temperature.

The light hitting you is heat flow in, which counteracts some of the heat flow out.

EDIT: Also, DSMat's explanation of quantum entanglement has now made it seem dull and pedestrian to me.

[Username17]

The "No Communication" theorem is frankly bullshit. Claiming that no "useful" information can be sent faster than the speed of light is not actually saving the no communication theorem. Just because you can't read a signal does not mean that a signal did not get sent and received. It just means you couldn't read it. But if the effect propagated faster than light, then no communication is already sunk.

As for the specifics of getting information you can use out of a quantum entanglement scenario, you do indeed need to have some sort of real information about the other side, which could entail having them send you a light speed radio broadcast to explain what they did afterward. But they could also send you a schedule of what they were going to do before hand. You wouldn't "know" that they actually did things correctly, because there would be no way to confirm for certain until afterward and in the Humian sense you can't ever know anything - but if you fucking trust the people you're talking to as far as accepting that they are going to continue operating the machine on the other end according to the established schedule for doing so - you're good. You can get enough information to harvest information out of a quantum entanglement by declaring ahead of time what that information is going to be and then having the other guys make that information.

Once there is ftl propagation, no communication is dead. Because regardless of what the people who suck causality cock try to convince themselves: garbled communication is still communication.

-Username17

[Quantumboost]

But correct me if I'm wrong, the outside temperature has an effect on radiation as well.

Nope. Three factors: the radiating body's temperature, the radiating body's surface area, and the tendency of the radiating body to radiate.

Technically, the outside temperature has an effect on radiative heat loss, in that radiative heat transfer from outside can dump heat back into the ship. But that's a negligible difference in space unless you're flying through a nebula or something, so it doesn't show up in the actual calculations at all.

[DSMatticus]

You can get enough information to harvest information out of a quantum entanglement by declaring ahead of time what that information is going to be and then having the other guys make that information.

This is the exact situation I've been describing and refuting. Your assertion: Alice can influence the probability distribution of Bob's measurements by choosing to measure or not. And this assertion is totally false. There is no measurement Bob can do that will tell him anything about what Alice has done or not. So even if Alice schedules, "I'm going to measure," Bob can't actually check if she measured or not later. Both events produce the same probability distribution. Bob can't verify that Alice is following the schedule, which means Bob doesn't know when she deviates from it. The only way to change this is to not only schedule measurements, but schedule results. And that's impossible: Alice can't schedule her results, because they're totally random.

Claiming that no "useful" information can be sent faster than the speed of light is not actually saving the no communication theorem.

This is just semantical. To physics, "information" means a very specific thing, and it generally reduces to anything that could be used to deduce a bit (or more). You're using a different definition than physicists. There is seriously a list of effects that already propagate faster than light. We've shown that these don't transmit information (in the physics sense). I can explain some of them, most of them I can't. Some of them look like they should be able to, but someone smarter than me knows why they don't.

[Username17]

DSM, that entire setup is based on a fringe interpretation that is desperately attempting to save causality by making shit up. Transmiting the state of a particle from one end of the world to the other is in no way similar to the apparent speed of far away objects while you are spinning. It's not even a little bit similar, because effects are actually happening at places that are distant and being affected in a real and measurably super luminal way.

The entire Alice and Bob thing you're talking about is not true because it's a claim being made to help defend a theory that was seriously killed more than a decade before I was even born. The idea that information can't go faster than light is dead. Even if it was actually true that we couldn't read that information, the theory would still be dead because it demonstrably has been sent.

I think it's cute that there are people like Hayden who are doing incredibly complicated math with temporary virtual universes to try to save the idea of local realism, but local realism was shown to be experimentally inconsistent with the actual universe in 1964. I don't fucking care what kind of beautiful math people come up with to prove local realism, because that shit is experimentally untrue. All an equation showing how local realism would work does is demonstrate that it is an equation that does not describe the real world we actually live in.

And that's the bottom line: you can make any math you want, but the universe is still the universe and it doesn't change its properties to match whatever string theory or multiple worlds hypothesis you happen to have. Local realism is false. Experimentally demonstrably false. And pretty math - no matter how pretty - cannot make it true.

-Username17

Edit: Zakharov said it best:
"A brave little theory, and actually quite coherent for a system of five or seven dimensions -- if only we lived in one."

[RadiantPhoenix]

I have a question: could someone explain how we actually know that information is sent when the particle is observed, as opposed to merely being already contained in the particle?

Let's take a metaphor:
* I have a sealed box
* my Martian pen-pal has a sealed box
* One of the boxes has one of these waveforms in it, and the other has an opposite waveform, but we don't know what they are yet.
* If I don't open my box, but my pen-pal does, he has a 50% chance of finding each waveform (which we will call X and Y)
* If I open my box first, I have a 50% chance of finding X or Y
* If I found X, my pen-pal has a 0% chance of finding X and a 100% chance of finding Y, and vice-versa.
* Also, if someone opens a box, the waveform in that box changes afterwards, so future observations of a box formerly containing X won't find X.

Obviously, there are some complications, but I think this covers the general gist of what I was getting from DSMat. It seems entirely unremarkable, so what am I missing?

EDIT: Also, I am not a quantum physicist.

[DSMatticus]

There are two distinctions you need to make: firstly, between locality and causality. Quantum entanglement requires either non-locality or hidden variables. The Bell Experiments disproved hidden variables, which means we invariably have non-locality. Now, locality is an axiomatic assumption of causality, so you would think that when locality goes causality would go too. Except no one has figured out a way to use non-locality to violate causality. Causality has experimentally survived the loss of local realism, one of its axioms.

It's possible that will change as we learn more, uncover new principles, or find weirder and weirder things to do with quantum entanglement. But what we have proven is that the specific type of thing you're talking about doesn't violate causality. We're certain of that.

The second distinction is this whole information thing:

Even if it was actually true that we couldn't read that information, the theory would still be dead because it demonstrably has been sent.

What quantum entanglement actually is is a pair of linked RNG's. The only way to extract a bit out your RNG would be to see "is it behaving any different than normal?" And what we've shown, with probability density functions that match the results provided by the Bell Experiments, is that the linked RNG's behave the same whether the other end is using their's or not. Ergo, you cannot extract a bit.

And you can't violate causality when you can't get a single bit of actual information.

I have a question: could someone explain how we actually know that information is sent when the particle is observed, as opposed to merely being already contained in the particle?

That theory is called the hidden variables theory; that the particle is storing its information and then going on a long journey. It turns out to be bullshit.

Remember those intermediate cases I was talking about, where you measure the x-axis and then you measure 10 degrees away from the x-axis? Bell did the math, and the hidden variables theory would imply that the correlation with respect to the difference in angle is linear. The Bell Experiments went on to show that the correlation is in fact not linear, it varies in a cosine-like wave. Ergo, hidden variables can suck it.

More specifically, the hidden variables theory predicts that when Alice measures the x-axis, and Bob measures 27.5 degrees away from the x-axis, their results should have 75% correlation. (Because 27.5 is 1/4th of the way between the x-axis and y-axis, which have 100% and 0% correlation: 1/4th of the way from 100% to 0% is 75%.) It turns out the correlation is experimentally ~86%, because it's at the peak of a cosine-wave. Hidden variables isn't consistent with experimental observation.

There really is an effect here that propagates instantaneously. It's just that we have found exactly zero ways to encode even one bit in this effect.

[tzor]

IIRC the best current star ship design (Project Orion) used nuclear explosions and that only got the ship to 10 psol (psol = percent of the speed of light). Even the effects of relativity are minor at this speed. Actually it would probably go much slower. The "momentum" limited model maintained a thrust of 1g for 10 days resulting in 3.3 psol.

If we upgrade to the good stuff ... "A nuclear pulse drive starship powered by matter-antimatter pulse units would be theoretically capable of obtaining a velocity between 50% to 80% of the speed of light."

IIRC at about 80 psol we start to see significant space time dilation effects. I forget the actual speed where it would take X years to travel "X light years" as measured from when you were originally stationary (so a trip to Alpha Centauri would be 4 "years" becase at your current speed space and time has so dilated that it's no longer 4 years away as the photon flies) but it was around that speed.

But you are still limited by the "human factor" of limiting long term acceleration to 1g.

[Username17]

What quantum entanglement actually is is a pair of linked RNG's. The only way to extract a bit out your RNG would be to see "is it behaving any different than normal?" And what we've shown, with probability density functions that match the results provided by the Bell Experiments, is that the linked RNG's behave the same whether the other end is using their's or not. Ergo, you cannot extract a bit.

And you can't violate causality when you can't get a single bit of actual information.

That is the no communication claim, but it's bullshit. If "all" it's telling you is the results of a meaningless random number simultaneously generated a light year away, that's still telling you something. It's real information. It may be stupid information that you don't give a fuck about, but it's real information and the no communication theorem is already busted.

The claim that you can't send a bit because of random variables being too noisy may or may not be true, but the theoretical framework the theory is created to defend is busted either way. If you receive information that a spin event on Alpha Centauri was one way or the other a minute ago, that's superluminal information even if you don't give a flying fuck which way it went.

-Username17

[Vebyast]

IIRC the best current star ship design (Project Orion) used nuclear explosions and that only got the ship to 10 psol (psol = percent of the speed of light).

Not quite; Project Orion is the best completely self-contained, non-megascale spaceship we've been able to come up with using only existing, available technology.

Two other very good designs:
[*] Fusion Candle (megascale): described in the author's note on this comic.
[*] Starwisp (non-self-contained): Built a gigantic solar-powered microwave laser in orbit around the sun. Throw a tiny payload out into space on the other end of a huge microwave sail. Boost the payload using your laser. Also works pretty well if you get invaded by aliens.

[Grek]

I'm sorry Frank, but no:

If we let A be Alice observing a positive spin and B be Bob observing the spin at all, then we apply Bayes Theorem, that P(A|B)=P(A)*P(B|A)/P(B). Since we know that Alice's actions a light year away aren't going to make Bob any more or less likely to follow the schedule and observe the particle at the proper time, P(B|A) = P(B) = P(B|~A) and therefore P(A) = 0.5 = P(~A)

P(A|B)=P(A)*P(B)/P(B)=0.5*1=0.5

When Alice tries to infer how Bob is doing based on the entangled particles they'd been planning to use to communicate:
P(B|A)=P(B)*P(A|B)/P(A)=P(B)*(0.5/0.5)=P(B)

And therefore Alice has no evidence about whether or not Bob has observed his particle(s) in an attempt to transmit a message. And therefore has no information from Bob.

[fectin]

Even intergalactic space has a temperature of between 2 and 3 Kelvin. That's the cosmic microwave background radiation.

I don't believe that external temperature is a factor in radiative heat dissipation rates. I'm pretty sure that's based on absolute and not relative temperature.

Rate of convective heat transfer across a material interface (e.g. hull to space) is the difference in temperatures times a constant for transfer between those two materials per unit area. Yes, that means it's a differential equation. For hard vacuum, that constant is going to be nearly zero (in fact, it will be zero, but you'll also boil off a few atoms of hull, and I'm not sure how that counts). That's why expensive thermoses are metal with vacuum inside. If the difference doubles, so does the transfer rate. That's why hotter things burn your skin worse/faster.
In space, convective transfer is going to be basically nil. You can do some funny stuff with decompression to get cooling, but that just shuffles your heat around.

You also have blackbody radiation, where a lump of stuff turns heat into photon emissions. It's generally a lot slower than convective heat transfer.

You can also eat energy with state changes and with chemical reactions.

Nuclear plants dump a lot of excess heat. I don't know how much, but it's a lot. Remember at Fukushima, they were bringing in a lot of water to dump on it? One of their issues was that it was hot enough to split the H2O apart, and create huge clouds of oxygen and hydrogen. That is a crapton of heat, and you need to be able to dump that somewhere, or you will very quickly boil.

See the earlier discussion on "can you cloak a spaceship".

[echoVanguard]

That is the no communication claim, but it's bullshit. If "all" it's telling you is the results of a meaningless random number simultaneously generated a light year away, that's still telling you something. It's real information. It may be stupid information that you don't give a fuck about, but it's real information and the no communication theorem is already busted.

The claim that you can't send a bit because of random variables being too noisy may or may not be true, but the theoretical framework the theory is created to defend is busted either way. If you receive information that a spin event on Alpha Centauri was one way or the other a minute ago, that's superluminal information even if you don't give a flying fuck which way it went.

-Username17

Weighing in on the side of DSMatticus and Grek here. To illustrate the principle in question, let's postulate a Schrodinger's Catphone. We have two scientists, separated by, oh, a light-year, each with a Schrodinger Box containing Cat(1) and a cyanide pill linked to the spin of a particle (as opposed to the decay of a radioactive atom, as in the standard Schrodinger Box design). Thus, if Cat 1 is dead, Cat 2 is alive, and vice versa, regardless of how much distance separate the two.

Scientist 1 opens the box, and discovers that his cat is alive - thus, he automatically knows that Cat 2 is now dead, despite the fact that the cat is a light-year away. However, this is of absolutely no use to anyone - Scientist 2 did not send a "my cat is dead" bit to Scientist 1, because Scientist 2 did not even open his box yet to discover that his cat has died. Similarly, Scientist 1 did not send a "my cat is alive" bit to Scientist 2, for the same reason. Finally (and this is key here) Scientist 1 has no way to distinguish a live cat from a dead cat until he opens the box, and he has no way of knowing if he was the first to open the box (and thus the sender or receiver of the catbit).

This holds true even if both scientists have magical spin-rays that manually control the state of the particle's spin, and are thus capable of setting the spin individually at will - although Scientist 1 can now send a "your cat is dead" bit to Scientist 2, Scientist 2 is completely incapable of distinguishing:
- opening his box and discovering a dead cat as a result of a "your cat is dead" bit from Scientist 1
from
- opening his box and discovering a dead cat even though Scientist 1 sent no information at all.

echo

[DSMatticus]

and he has no way of knowing if he was the first to open the box (and thus the sender or receiver of the catbit).

Not quite what Frank's saying. Frank has claimed (and is correct) that you can circumvent this particular problem by scheduling the measurements. I.e., scientist 1 and 2 are light-years away, and scientist 2 is scheduled to measure his particle one minute later than scientist 1. Ergo, as soon as scientist 2 opens his box he learns about an event that happened a light year away even though it happened only a minute ago.

But this sort of thing happens all the time, and it's not sufficient to show FTL transfer of information. Here's a classical example:
1) You cut a coin down the middle, producing a heads-side and a tails-side. You randomly put each in an envelope.
2) You hand these envelopes to two individuals, and tell them travel a few light years apart. You schedule person A to open their envelop one minute before person B.
3) Person A opens their envelope, and gets some result.
4) Person B opens their envelope, and gets the opposite result. They now also know person A's result, even though they are light-years away and the event happened a minute ago.
C) Was this FTL transfer of information? No. You're deducing a third piece of information from two others: if you get result X (subluminal information, personally observed), and their result is the complement of your's (subluminal information, known from the experiment description), ergo their result was the opposite of X (deduced, not transmitted).

The reason the quantum mechanical equivalent throws everyone for a loop is because the values in the coin experiment are decided at the beginning of the experiment, and the values in the quantum experiment are decided the first time someone measures. It fucks with human intuition, and it does require the propagation of some effect (but not necessarily information) FTL. But the lesson to take away holds whether it's classical coins or quantum particles: "spooky learning at a distance" is necessary but not sufficient to show FTL communication. We need something more to show a transfer of information occurred faster-than-light.

(But even with scheduling, the problem I've been trying to describe holds. No action you can take effects the probability of results experienced at the other end, so you can't send a bit. Grek did the explicit math for the 1-dimensional case where Bob and Alice are both measuring the same axis. The math is a lot more complicated as you add dimensions and the ability to measure different axes, but the results don't change.)

EDIT: P.S...
magical spin rays + scheduling -> FTL communication. Scientist 1 can force either +X or -X to transmit bits, and then inbetween those intervals scientist 2 measures his particle to receive them.

persistent entangling + scheduling -> FTL communication. Scientist 1 measures his particle until he gets the desired result (the bigger the gap in the schedule, the more time he has to get the desired result; it works out to be equivalent to forcing with a transmission error rate that approaches 0 as time approaches infinity).

[Doom]

I'm sorry Frank, but no:

If we let A be Alice observing a positive spin and B be Bob observing the spin at all, then we apply Bayes Theorem, that P(A|B)=P(A)*P(B|A)/P(B). Since we know that Alice's actions a light year away aren't going to make Bob any more or less likely to follow the schedule and observe the particle at the proper time, P(B|A) = P(B) = P(B|~A) and therefore P(A) = 0.5 = P(~A)

Your "Alice's actions..." aren't going to affect Bob's actions line is just assuming independence. You're assuming independence, and then concluding from this that there's independence. This is meaningless, and you certainly can't even conclude P(A) = 0.5.

Under the assumption A and B are independent, we have P(A|B)= P(A), and P(B|A) = P(B).

Bayes theorem is irrelevant in this case, as all it asserts is that
P(A) = P(A), and nothing more.

[Grek]

I presume that you're not familiar with the idea of a bayesian update of beliefs?

With bayesian statistics, you start out with a prior probability based on all of your prior knowledge of the topic at hand. Then, whenever you see evidence, you update your beliefs based on your prior probability and a likelihood ratio for the evidence to arrive at a posterior probability. Which then serves as your prior for the next bit of evidence you consider.

In this case, that likelihood ratio is 1:1, so according to the bayesian definition of evidence, Alice has no reason to update her beliefs about Bob and Bob has no way of giving Alice evidence that would make it reasonable for her to update her beliefs. And unless someone's changing their mind and learning new knowledge, no communication can properly be said to have occured.