P.S.: maybe it will help clarify things to you to read the Wikipedia article on Spacetime, and to note that another common term (when discussing spacetime) for what I'm calling 'points' is 'events'. From the Wikipedia article:
an event is a point in spacetime specified by its time and place.
Because events are spacetime points, an example of an event in classical relativistic physics is (x,y,z,t), the location of an elementary (point-like) particle at a particular time. A spacetime itself can be viewed as the union of all events in the same way that a line is the union of all of its points, formally organized into a manifold, a space which can be described at small scales using coordinates systems.
In a Euclidean space, the separation between two points is measured by the distance between the two points. A distance is purely spatial, and is always positive. In spacetime, the separation between two events is measured by the invariant interval between the two events, which takes into account not only the spatial separation between the events, but also their temporal separation.
The causal structure of a spacetime describes causal relationships between pairs of points in the spacetime based on the existence of certain types of curves joining the points.
??? Do you have problems with the idea of a "point" in Cartesian coordinates, either? A problem with the concept of "5pm at the corner of Main and Central" defining a spacetime point independent of the motion of Main and Central and what happens at any time other than 5pm? Do you ask questions like "where is the point (x,y)=(1,3) when y=1.8"? An observer is "someone" who follows a path through space-time (their "world line"); along this path, you can measure things like velocity, and whether said observer is "co-moving" or not. Like any other curve, the path is an (infinite) set of points in space-time, just like a points along a line in a Cartesian space (which don't have a "slope" even through the line does). Take the most introductory course on special relativity and you'll be talking about spacetime points right at the beginning --- they are quite well defined, and critical to basic concepts like the "invariant interval" between two points ds^2 = (space distance)^2-c^2*(time difference)^2, which remains the same no matter what frame you observe from (though the space and time separations between two points vary depending on your reference frame, as do coordinates that you might use to label said points).
You could surely just set up a scheme where by if you cause a state change in 1_1 and 1_2 simultaneously, you can also detect that 2_1 and 2_2 changed simultaneously
No, you surely could not (well, unless you have some radical new theory to replace quantum mechanics, and said theory is correct in its relevant predictions). There is no way to tell when a state is changing, whether that's a single-particle state or any number of particle states. No fancier encoding scheme gets you around this.
Again, you don't understand the statements I'm making. There is no such thing as "co-moving A and B": observers can be co-moving; it makes no sense for space time *points* (a location through which an observer can move). Suppose you have a pair of black boxes that, when sitting still in your reference frame, can send FTL messages in *your* reference frame (one box sends a message at space,time point A, and the other receives it at space,time point B, which don't have a velocity of their own, but in your frame A precedes B). Then you can use these exact same pair of boxes to send a message *backwards* in time in *your* rest frame. In order to do that, you'll need to throw the boxes so they're moving fast relative to you --- in their own reference frame, they're still sending a forward-in-time message. However, in your reference frame, the "receiver" box whizzing through one lab will be able to get a message from the "sender" box whizzing through a different lab *LATER*. By chaining a couple of these together, you can relay a message back to your past self to tell your future self not to send the message.
In my example, "A" and "B" refer to points in space time. They have no velocity of their own; there's no meaning to "A moving fast relative to B", or "moving fast relative to A" ("5pm at the corner of Central and Main" doesn't have a velocity, even though "the corner of Central and Main" does). Using somewhat loose language, I also used them to refer to an observer whose world-line passes through point A or B --- someone who is at a particular place at a particular time ("A" or "B"), but could be moving at any velocity passing through said point. Depending on your reference frame, A and B may be separated differently in time and space (and do not themselves depend on choice of frame). My argument shows causality violation if you have FTL transmission in any reference frame, since there is some other reference frame where the message is received before being sent.
The "small change in probability" I was referring to was the probability distribution for, e.g., polarizations on the "sender" side that he's manipulating. It's easy to make a small, or extremely large, change in probabilities on your side (you can re-polarize 100% of your electrons to be "spin up" if you want, but you'll 100% break entanglement in the process ; trying to slip past smaller changes doesn't help either). It is indeed accurate to say there is "zero corresponding change" on the other side, which was the point of my post.
If you absolutely trust SSL v2 to be un-crackable, then there's no "point" to better communications. However, if there is the least bit of worry that classical cryptography can be broken (either due to flaws in the algorithm that a super-smart mathematician might discover, or "brute-force" solutions by quantum computers --- the jury is still out whether such is possible), then quantum transmission channels offer a new "layer" of security not vulnerable to clever mathematicians or super-powerful (quantum) computers.
From another perspective, even if quantum entanglement has no commercial application, it's still cool to learn how the universe works simply for the sake of curiosity.
Information is basically the "minimum" stuff that needs to be sent to have "causality". For example, I can cause you to be or not be punched in the face by either walking over and punching you in the face or not; but that requires a lot of work on my side. What's the "minimum" I have to do? Well, perhaps you're already sitting in front of a face-punching machine. All I need to be able to do to have "causal" impact on your face is send some "information" to the face-puncher machine; a tiny electrical pulse to the "start punch!" circuit. If I can't even send "information," then I can't do anything at all; I'm no longer causally connected to your face getting punched.
Thus, the ability to send information (hence meet the minimum requirements for causality) is very important. If not even information can be sent between two regions, then they are not in causal contact. Some things can "move faster than c," so long as they carry no causality-relevant content (i.e. no information). Transmitting information, however, creates causal contact --- and any causal contact faster than c permits causal paradoxes, where the effect of a cause can precede the cause (!), which, so far as we know, can't happen (and doesn't make "sense" if it could, since "sense" is based on non-paradoxical causality).
The problem is your forms of manipulation on the "non-entangled" partner don't "transmit" to the entangled partner --- no matter what you do, it'll still be 50/50. To the extent that you "partly" measure the polarization and gently tweak the distribution of particles on your end, you also "partly" break the entanglement: there is no longer 100% correlation. The more you "coerce" your particles into a better-known state, the weaker the entanglement correlation becomes, in proportion such that you *never* get even 1% transmission of information. Like a lot of "perpetual motion machine" plans, you're trying to skim a small change in probability on one side (saying it's big enough to add up to something useful), while ignoring the corresponding small changes in correlation on the other (assuming they're negligible because they're so small).
First, for pedantic clarity, no "information" is transmitted through acts of fiddling with entangled bits. The "information" is transmitted as the entangled particles are moved to their respective endpoints (at no more than the speed of light), and though further actions may "change the outcome" (the exact meaning of which requires rather subtle interpretation, and can be framed within many not-incompatible understandings of QM), they cannot "transmit information".
Your question, then, is no different in the "quantum" case than for any classical information transfer. If I mail you a letter, has any information been transmitted if you don't verify with me the integrity of the information in the received letter? How do you verify the integrity of the information, except through a transfer of information that itself requires verification? If you make the basic assumption that science relies on, that the universe operates according to repeatable principles, then you can achieve at least a statistical certainty that information is properly transmitted without an endless chain of uncertainty ("this method worked the 99 times when we checked before, so I bet this 100th transmission also conveyed information").
Since "the universe as simulation" is an un-disprovable proposition that could be made regardless of what we observe (really, what would indicate a non-simulated universe? flaming letters in the sky saying "THIS IS NOT A TEST"?), folks use all sorts of "evidence" to support it (especially "wacky" quantum concepts poorly understood by the general public and some metaphysical philosophers). This doesn't prove the universe isn't a simulation, either; only that, so long as the proposition remains untestable (we can't "break through" to observe the "outer layer"), such metaphysical arguments about the universe are entirely useless within the realm of science.
Your ISS partner doesn't have a '1' filter. They have an "all 1's become zeros, all 0's become 1's" filter. In both cases, you just see a random stream of zeros and ones. As soon as your ISS partner "looks" at a bit to tell whether its a 0 or 1 (e.g., if they want to apply a filter only to the 1's), they break the entanglement on that bit --- after that point, they can flip the bit all they want and it does nothing on your end (you just see the random bit that's opposite of whatever they measured at the point of breaking disentanglement).
Sorry, stupid slip-up (need more coffee!) --- there's no reference frame where A and B are at the same location and reverse time order. However, you can pass through a frame where A and B are closer together and at the same time, to where A and B are further apart but in reverse time order (which you can't do for points that are closer in space than in time in any frame): FTL communications forwards-in-time would also allow communications backwards-in-time if the sending/receiving devices where thrown at the right velocity across the lab, and from there you can build all sorts of paradoxes using causality violation.
In a sense, you can't "tell" the switching machine works on particles you aren't allowed to measure on both sides --- for the same reason that you can't "tell" there isn't a giraffe that appears in your closet whenever the door is closed and you can't look. However, the switching machine can be one that's proven to work whenever you pass a known-state particle through (just like you can prove there's no giraffe in every instance that you open the closet door); you can make a polarization rotator that works every single time you check a particle's state before and after. Furthermore, the observable data that you measure is never inconsistent with the polarization also working on the unobserved particles. Thus, it's sensible to think about the universe in terms of the polarization-rotator always working (instead of magically "breaking" just and only for the cases you don't check), just like it's sensible to operate with a no-giraffes-in-closet theory (even though an unobservable-giraffe-in-closet-only-when-you-don't-check theory is just as good a description of known data).
Here's how FTL communication can break causality. In your reference frame, spacetime points A and B are 2 light years apart, with B "happening" 1 year after A. A sends an FTL message to B (arriving 2 light years away, but 1 year later).
Here's where relativity messes you up: by changing your reference frame, you can trade off between the "space" and "time" differences between A and B, so long as the "invariant interval" s^2 =(space difference)^2 - c^2*(time difference)^2 remains constant, and continuously connected on the same branch of the hyperbola s^2 = constant (in this case, s^2 = 2^2-1^2 = 3). In particular, you can boost through a reference frame where (space difference) = sqrt(3), (time difference) = 0, continuously to a point where (space difference) = 2, (time difference) = -1. In this frame, A and B are the same difference apart, but occur in the opposite time order (B a year before A!). In the case of space-difference s^2 = dl^2 > 0 (contradiction 0>0).
Thus, in the alternate frame, B receives a FTL message a year before A sends it! The same technology that allows A to send a FTL message "forward in time" in one frame allows messaging "backwards in time" in another. This allows construction of paradoxical triangles: B sends a message to A (2 LY away, a year earlier); A sends a reply to B' (in the same location but 2 years earlier than B) saying "don't send your message"; B' leaves a post-it on the lab wall saying "two years from now, don't send a message back to A".
Another example of "classical superluminal" properties is sweeping a laser pointer across a distant wall: there's no classical physical law which prevents you from rotating a laser pointer so the point of light on a distant wall moves from point A to B "faster than the speed of light." In this case also, there's no way to transmit information from A to B faster than light via this method. Your linked paper is a clever fancier version of this --- setting up "extended systems," so it appears that a polarization ripple is moving across a system FTL (like a bright point moving FTL across a wall), due to slower-than-light "preparation" of the system (like sending initial photons from the laser in the right directions to produce the "FTL" spot). Quantum "FTL" effects may indeed fall into the same "philosophical" category of "apparently-FTL due to clever preparation of the system" (initial production and non-FTL transport of the entangled particles).
FTL communication interferes with causality (within Einsteinian relativity, which so far seems to be a pretty solid theory) because if you have something that looks like FTL communication in one reference frame (a signal is sent from spacetime point A which arrives 1 year later at spacetime point B, which is 2 light years away from A), then there exists some other frame in which causality is violated: B receives the signal *before* A sends it (in this case, there's another frame where A and B are at the *same location,* except B is the time *a year before* A sends the signal; this would allow all sorts of causality-violating paradoxes, like B leaving a message for A saying "don't send the message we just received from you!"). You may consider the circumstances "contrived" in the sense that you need huge velocity differences between the frame in which A precedes B, and that in which B precedes A --- however, "contrived" or not, the theory breaks down. A theory describing how the universe works shouldn't depend on how good your ansible structural engineers are; and, we actually can create particles traveling much faster than the required frame boost for this example, and observe that they behave according to relativistic theory.
Nope. You can't tell on the receiving end whether or not the bit you're measuring was or was not rotated. A measured '1' could be a 'rotated 0' or an 'unrotated 1', so you know absolutely nothing about what the message sender is doing (only that, if she measured her bits right after you measured yours, she'd see the opposite values).
Doesn't work --- on the receiving end, you have no way to tell whether or not the bit was rotated on the opposite side. You see a random bit; suppose it's a 1. There's no way to tell whether this was a 0 "to start with" then rotated, or a 1 "unrotated." The original states of the bits are random: you can't produce entangled bits knowing that one side starts with all 1's, and the other starts with all 0's.
I am a professional physicist. Entanglement does indeed appear to be "instantaneous over long distances". However, no data is transmitted through this process. You're half-right, half-wrong: right on the part most people don't argue against, and entirely wrong on the important question of data transmission. And you're all wrong on thinking you're in any way qualified to comment on this question, because you're obviously not.
Perhaps there was a Slashdot article with a lousy summary (shocking!) that mislead uninformed people into thinking NASA had a method for transmitting data between entangled particles, but NASA does not have a method for transmitting data between entangled particles.
The problem in your scheme is "set one set of pairs to some value": you can't do that. Quantum mechanics forbids you from setting one side of the pair to values you want (without first breaking the entanglement, so the other side won't see the results).
Depending on the experiment, 10% differences can be pretty obvious to measure. With the best atomic clocks, we can now see relativistic effects due to gravitational potential differences corresponding to 1m height change in the lab. Without understanding the experiment, you have no way to judge whether 10% differences are negligible or whoppingly huge compared to experimental sensitivity.
Up to the point you describe, the classical (a pair of cards with different colors) and quantum (a pair of photons with different polarizations) systems behave the same. However, there are some nifty features of the quantum system that can't happen classically. For example, you can rotate the polarization of one entangled photon (without knowing what it is) --- then, measure both photons, and they'll still be in opposite polarizations, even if you carried out the polarization rotation and measurement so far apart that no "signal" had time to travel, even at the speed of light, from one to the other (what Einstein would call "spooky action at a distance").
Because the universe doesn't seem to like causation violation, so all its operating principles preclude faster-than-light (which, in Einsteinian relativity, is equivalent to "faster-than-causality") information transmission.
A rough "classical" analogy for quantum entanglement is: seal two cards, one white and one black, in a pair of envelopes. Shuffle the envelopes, and give one to a person who travels to the Moon. Whenever they open their envelope, they'll instantaneously know what the other envelope contains. However, this doesn't instantaneously "transmit" any information: all the information was "transmitted" when the person carried their envelope to the moon, at under the speed of light.
The "quantum" part of Quantum Entanglement adds some fun not-in-classical-physics features to this analogy. For example, you can make a machine that will flip a black card to white and white to black (without telling you which); when the person on the moon puts their envelope through such a device, it can still stay "in sync" with the other envelope (when they are both opened afterwards, they'll still have opposite-colored cards). However, no information is transmitted: the Earth person has no way of knowing (unless you tell them through speed-of-light-or-slower channels) whether or not the Moon person has used the card-flipping machine; once they've checked their own envelope, the entanglement is broken and changing the Moon envelope's contents no longer changes the one on Earth.
Slashdot Mirror
P.S.: maybe it will help clarify things to you to read the Wikipedia article on Spacetime, and to note that another common term (when discussing spacetime) for what I'm calling 'points' is 'events'. From the Wikipedia article:
an event is a point in spacetime specified by its time and place.
Because events are spacetime points, an example of an event in classical relativistic physics is (x,y,z,t), the location of an elementary (point-like) particle at a particular time. A spacetime itself can be viewed as the union of all events in the same way that a line is the union of all of its points, formally organized into a manifold, a space which can be described at small scales using coordinates systems.
In a Euclidean space, the separation between two points is measured by the distance between the two points. A distance is purely spatial, and is always positive. In spacetime, the separation between two events is measured by the invariant interval between the two events, which takes into account not only the spatial separation between the events, but also their temporal separation.
The causal structure of a spacetime describes causal relationships between pairs of points in the spacetime based on the existence of certain types of curves joining the points.
??? Do you have problems with the idea of a "point" in Cartesian coordinates, either? A problem with the concept of "5pm at the corner of Main and Central" defining a spacetime point independent of the motion of Main and Central and what happens at any time other than 5pm? Do you ask questions like "where is the point (x,y)=(1,3) when y=1.8"? An observer is "someone" who follows a path through space-time (their "world line"); along this path, you can measure things like velocity, and whether said observer is "co-moving" or not. Like any other curve, the path is an (infinite) set of points in space-time, just like a points along a line in a Cartesian space (which don't have a "slope" even through the line does). Take the most introductory course on special relativity and you'll be talking about spacetime points right at the beginning --- they are quite well defined, and critical to basic concepts like the "invariant interval" between two points ds^2 = (space distance)^2-c^2*(time difference)^2, which remains the same no matter what frame you observe from (though the space and time separations between two points vary depending on your reference frame, as do coordinates that you might use to label said points).
You could surely just set up a scheme where by if you cause a state change in 1_1 and 1_2 simultaneously, you can also detect that 2_1 and 2_2 changed simultaneously
No, you surely could not (well, unless you have some radical new theory to replace quantum mechanics, and said theory is correct in its relevant predictions). There is no way to tell when a state is changing, whether that's a single-particle state or any number of particle states. No fancier encoding scheme gets you around this.
Again, you don't understand the statements I'm making. There is no such thing as "co-moving A and B": observers can be co-moving; it makes no sense for space time *points* (a location through which an observer can move). Suppose you have a pair of black boxes that, when sitting still in your reference frame, can send FTL messages in *your* reference frame (one box sends a message at space,time point A, and the other receives it at space,time point B, which don't have a velocity of their own, but in your frame A precedes B). Then you can use these exact same pair of boxes to send a message *backwards* in time in *your* rest frame. In order to do that, you'll need to throw the boxes so they're moving fast relative to you --- in their own reference frame, they're still sending a forward-in-time message. However, in your reference frame, the "receiver" box whizzing through one lab will be able to get a message from the "sender" box whizzing through a different lab *LATER*. By chaining a couple of these together, you can relay a message back to your past self to tell your future self not to send the message.
In my example, "A" and "B" refer to points in space time. They have no velocity of their own; there's no meaning to "A moving fast relative to B", or "moving fast relative to A" ("5pm at the corner of Central and Main" doesn't have a velocity, even though "the corner of Central and Main" does). Using somewhat loose language, I also used them to refer to an observer whose world-line passes through point A or B --- someone who is at a particular place at a particular time ("A" or "B"), but could be moving at any velocity passing through said point. Depending on your reference frame, A and B may be separated differently in time and space (and do not themselves depend on choice of frame). My argument shows causality violation if you have FTL transmission in any reference frame, since there is some other reference frame where the message is received before being sent.
The "small change in probability" I was referring to was the probability distribution for, e.g., polarizations on the "sender" side that he's manipulating. It's easy to make a small, or extremely large, change in probabilities on your side (you can re-polarize 100% of your electrons to be "spin up" if you want, but you'll 100% break entanglement in the process ; trying to slip past smaller changes doesn't help either). It is indeed accurate to say there is "zero corresponding change" on the other side, which was the point of my post.
If you absolutely trust SSL v2 to be un-crackable, then there's no "point" to better communications. However, if there is the least bit of worry that classical cryptography can be broken (either due to flaws in the algorithm that a super-smart mathematician might discover, or "brute-force" solutions by quantum computers --- the jury is still out whether such is possible), then quantum transmission channels offer a new "layer" of security not vulnerable to clever mathematicians or super-powerful (quantum) computers.
From another perspective, even if quantum entanglement has no commercial application, it's still cool to learn how the universe works simply for the sake of curiosity.
Information is basically the "minimum" stuff that needs to be sent to have "causality". For example, I can cause you to be or not be punched in the face by either walking over and punching you in the face or not; but that requires a lot of work on my side. What's the "minimum" I have to do? Well, perhaps you're already sitting in front of a face-punching machine. All I need to be able to do to have "causal" impact on your face is send some "information" to the face-puncher machine; a tiny electrical pulse to the "start punch!" circuit. If I can't even send "information," then I can't do anything at all; I'm no longer causally connected to your face getting punched.
Thus, the ability to send information (hence meet the minimum requirements for causality) is very important. If not even information can be sent between two regions, then they are not in causal contact. Some things can "move faster than c," so long as they carry no causality-relevant content (i.e. no information). Transmitting information, however, creates causal contact --- and any causal contact faster than c permits causal paradoxes, where the effect of a cause can precede the cause (!), which, so far as we know, can't happen (and doesn't make "sense" if it could, since "sense" is based on non-paradoxical causality).
The problem is your forms of manipulation on the "non-entangled" partner don't "transmit" to the entangled partner --- no matter what you do, it'll still be 50/50. To the extent that you "partly" measure the polarization and gently tweak the distribution of particles on your end, you also "partly" break the entanglement: there is no longer 100% correlation. The more you "coerce" your particles into a better-known state, the weaker the entanglement correlation becomes, in proportion such that you *never* get even 1% transmission of information. Like a lot of "perpetual motion machine" plans, you're trying to skim a small change in probability on one side (saying it's big enough to add up to something useful), while ignoring the corresponding small changes in correlation on the other (assuming they're negligible because they're so small).
First, for pedantic clarity, no "information" is transmitted through acts of fiddling with entangled bits. The "information" is transmitted as the entangled particles are moved to their respective endpoints (at no more than the speed of light), and though further actions may "change the outcome" (the exact meaning of which requires rather subtle interpretation, and can be framed within many not-incompatible understandings of QM), they cannot "transmit information".
Your question, then, is no different in the "quantum" case than for any classical information transfer. If I mail you a letter, has any information been transmitted if you don't verify with me the integrity of the information in the received letter? How do you verify the integrity of the information, except through a transfer of information that itself requires verification? If you make the basic assumption that science relies on, that the universe operates according to repeatable principles, then you can achieve at least a statistical certainty that information is properly transmitted without an endless chain of uncertainty ("this method worked the 99 times when we checked before, so I bet this 100th transmission also conveyed information").
Since "the universe as simulation" is an un-disprovable proposition that could be made regardless of what we observe (really, what would indicate a non-simulated universe? flaming letters in the sky saying "THIS IS NOT A TEST"?), folks use all sorts of "evidence" to support it (especially "wacky" quantum concepts poorly understood by the general public and some metaphysical philosophers). This doesn't prove the universe isn't a simulation, either; only that, so long as the proposition remains untestable (we can't "break through" to observe the "outer layer"), such metaphysical arguments about the universe are entirely useless within the realm of science.
Your ISS partner doesn't have a '1' filter. They have an "all 1's become zeros, all 0's become 1's" filter. In both cases, you just see a random stream of zeros and ones. As soon as your ISS partner "looks" at a bit to tell whether its a 0 or 1 (e.g., if they want to apply a filter only to the 1's), they break the entanglement on that bit --- after that point, they can flip the bit all they want and it does nothing on your end (you just see the random bit that's opposite of whatever they measured at the point of breaking disentanglement).
Sorry, stupid slip-up (need more coffee!) --- there's no reference frame where A and B are at the same location and reverse time order. However, you can pass through a frame where A and B are closer together and at the same time, to where A and B are further apart but in reverse time order (which you can't do for points that are closer in space than in time in any frame): FTL communications forwards-in-time would also allow communications backwards-in-time if the sending/receiving devices where thrown at the right velocity across the lab, and from there you can build all sorts of paradoxes using causality violation.
In a sense, you can't "tell" the switching machine works on particles you aren't allowed to measure on both sides --- for the same reason that you can't "tell" there isn't a giraffe that appears in your closet whenever the door is closed and you can't look. However, the switching machine can be one that's proven to work whenever you pass a known-state particle through (just like you can prove there's no giraffe in every instance that you open the closet door); you can make a polarization rotator that works every single time you check a particle's state before and after. Furthermore, the observable data that you measure is never inconsistent with the polarization also working on the unobserved particles. Thus, it's sensible to think about the universe in terms of the polarization-rotator always working (instead of magically "breaking" just and only for the cases you don't check), just like it's sensible to operate with a no-giraffes-in-closet theory (even though an unobservable-giraffe-in-closet-only-when-you-don't-check theory is just as good a description of known data).
Here's how FTL communication can break causality.
In your reference frame, spacetime points A and B are 2 light years apart, with B "happening" 1 year after A.
A sends an FTL message to B (arriving 2 light years away, but 1 year later).
Here's where relativity messes you up: by changing your reference frame, you can trade off between the "space" and "time" differences between A and B, so long as the "invariant interval" s^2 =(space difference)^2 - c^2*(time difference)^2 remains constant, and continuously connected on the same branch of the hyperbola s^2 = constant (in this case, s^2 = 2^2-1^2 = 3). In particular, you can boost through a reference frame where (space difference) = sqrt(3), (time difference) = 0, continuously to a point where (space difference) = 2, (time difference) = -1. In this frame, A and B are the same difference apart, but occur in the opposite time order (B a year before A!). In the case of space-difference s^2 = dl^2 > 0 (contradiction 0>0).
Thus, in the alternate frame, B receives a FTL message a year before A sends it! The same technology that allows A to send a FTL message "forward in time" in one frame allows messaging "backwards in time" in another. This allows construction of paradoxical triangles: B sends a message to A (2 LY away, a year earlier); A sends a reply to B' (in the same location but 2 years earlier than B) saying "don't send your message"; B' leaves a post-it on the lab wall saying "two years from now, don't send a message back to A".
Another example of "classical superluminal" properties is sweeping a laser pointer across a distant wall: there's no classical physical law which prevents you from rotating a laser pointer so the point of light on a distant wall moves from point A to B "faster than the speed of light." In this case also, there's no way to transmit information from A to B faster than light via this method. Your linked paper is a clever fancier version of this --- setting up "extended systems," so it appears that a polarization ripple is moving across a system FTL (like a bright point moving FTL across a wall), due to slower-than-light "preparation" of the system (like sending initial photons from the laser in the right directions to produce the "FTL" spot). Quantum "FTL" effects may indeed fall into the same "philosophical" category of "apparently-FTL due to clever preparation of the system" (initial production and non-FTL transport of the entangled particles).
FTL communication interferes with causality (within Einsteinian relativity, which so far seems to be a pretty solid theory) because if you have something that looks like FTL communication in one reference frame (a signal is sent from spacetime point A which arrives 1 year later at spacetime point B, which is 2 light years away from A), then there exists some other frame in which causality is violated: B receives the signal *before* A sends it (in this case, there's another frame where A and B are at the *same location,* except B is the time *a year before* A sends the signal; this would allow all sorts of causality-violating paradoxes, like B leaving a message for A saying "don't send the message we just received from you!"). You may consider the circumstances "contrived" in the sense that you need huge velocity differences between the frame in which A precedes B, and that in which B precedes A --- however, "contrived" or not, the theory breaks down. A theory describing how the universe works shouldn't depend on how good your ansible structural engineers are; and, we actually can create particles traveling much faster than the required frame boost for this example, and observe that they behave according to relativistic theory.
Nope. You can't tell on the receiving end whether or not the bit you're measuring was or was not rotated. A measured '1' could be a 'rotated 0' or an 'unrotated 1', so you know absolutely nothing about what the message sender is doing (only that, if she measured her bits right after you measured yours, she'd see the opposite values).
Doesn't work --- on the receiving end, you have no way to tell whether or not the bit was rotated on the opposite side. You see a random bit; suppose it's a 1. There's no way to tell whether this was a 0 "to start with" then rotated, or a 1 "unrotated." The original states of the bits are random: you can't produce entangled bits knowing that one side starts with all 1's, and the other starts with all 0's.
I am a professional physicist. Entanglement does indeed appear to be "instantaneous over long distances". However, no data is transmitted through this process. You're half-right, half-wrong: right on the part most people don't argue against, and entirely wrong on the important question of data transmission. And you're all wrong on thinking you're in any way qualified to comment on this question, because you're obviously not.
Perhaps there was a Slashdot article with a lousy summary (shocking!) that mislead uninformed people into thinking NASA had a method for transmitting data between entangled particles, but NASA does not have a method for transmitting data between entangled particles.
The problem in your scheme is "set one set of pairs to some value": you can't do that. Quantum mechanics forbids you from setting one side of the pair to values you want (without first breaking the entanglement, so the other side won't see the results).
Depending on the experiment, 10% differences can be pretty obvious to measure. With the best atomic clocks, we can now see relativistic effects due to gravitational potential differences corresponding to 1m height change in the lab. Without understanding the experiment, you have no way to judge whether 10% differences are negligible or whoppingly huge compared to experimental sensitivity.
Up to the point you describe, the classical (a pair of cards with different colors) and quantum (a pair of photons with different polarizations) systems behave the same. However, there are some nifty features of the quantum system that can't happen classically. For example, you can rotate the polarization of one entangled photon (without knowing what it is) --- then, measure both photons, and they'll still be in opposite polarizations, even if you carried out the polarization rotation and measurement so far apart that no "signal" had time to travel, even at the speed of light, from one to the other (what Einstein would call "spooky action at a distance").
Because the universe doesn't seem to like causation violation, so all its operating principles preclude faster-than-light (which, in Einsteinian relativity, is equivalent to "faster-than-causality") information transmission.
A rough "classical" analogy for quantum entanglement is: seal two cards, one white and one black, in a pair of envelopes. Shuffle the envelopes, and give one to a person who travels to the Moon. Whenever they open their envelope, they'll instantaneously know what the other envelope contains. However, this doesn't instantaneously "transmit" any information: all the information was "transmitted" when the person carried their envelope to the moon, at under the speed of light.
The "quantum" part of Quantum Entanglement adds some fun not-in-classical-physics features to this analogy. For example, you can make a machine that will flip a black card to white and white to black (without telling you which); when the person on the moon puts their envelope through such a device, it can still stay "in sync" with the other envelope (when they are both opened afterwards, they'll still have opposite-colored cards). However, no information is transmitted: the Earth person has no way of knowing (unless you tell them through speed-of-light-or-slower channels) whether or not the Moon person has used the card-flipping machine; once they've checked their own envelope, the entanglement is broken and changing the Moon envelope's contents no longer changes the one on Earth.