How accurate is the KSP interstellar Alcubierre Drive?

[Z-Man]

Yeah... that. I'm not even sure what Ben is trying to say.

The first triangle also does not exist. It is a mathematical construct.

Copenhagen looses the superposition at every point, and fails to keep it alive, but allows for an interaction. MWI keeps the superposition alive at evey point, but does not allow for an interaction.

No. Copenhagen prescribes a collapse after every measurement, not every point. There has been a lot of calm and rational (cough) debate about what a measurement is, but if done right, it loses nothing, it just transforms a superposition that can still interfere into a classical probability distribution which can't. And the transition happens when interference would become practically impossible.

And MWI does allow for interference (is that what you mean by interaction? It better be) between the branches. It's very important early on in the life of new branches. Only later, interference gets increasingly unlikely.

I really don't get where you get the idea from that any of these are inadequate tools for building models. You should probably look up at least a bit of the actual math involved and how it is applied to simple systems such as the spin of a single electron.

Edited November 26, 2014 by Z-Man

[Technical Ben]

Thanks. So I have a question on it then, and it's back on topic. How can MWI allow both time travel and not allow detection of alternate "worlds" (of what ever type we define)?

Either we can have detection, or we cannot. So where or what would we time travel into, if not an alternate universe?

That's not arm chair philosophy (it never was), it's asking a question on what I've seen as a contradiction being presented.

PS, Copenhagen cannot define a "measurement" and MWI cannot define a "split". It's the age old problem of two equal, but incomplete, solutions to a problem. Why argue over them?

I meant that MWI does not allow for any event to happen. The basic explanation given was that MWI allows for no "objects", it has no collapses (I'm unsure if this is limited to a time like interaction or not). Is that correct?

Edited November 26, 2014 by Technical Ben

[Rakaydos]

Ok, I have a question on it then, and it's back on topic. How can MWI allow both time travel and not allow detection of alternate "worlds" (of what ever type we define)?

Either we can have detection, or we cannot. So where or what would we time travel into, if not our own universe?

That's not arm chair philosophy (it never was), it's asking a question on what I've seen as a contradiction being presented.

[Technical Ben]

That's not time travelling.

That's spontaneous "last Tuesdayism".

AFAIK Copenhagen does not forbid that either. We could also detect this as different from time travel, as we would see the dinosaurs being ridden around the streets in the alternate universe.

So as far as I can see, either MWI does not allow time travel and does not allow other branch detection/interaction or it allows both. Or is there a special case?

[Z-Man]

Thanks. So I have a question on it then, and it's back on topic. How can MWI allow both time travel and not allow detection of alternate "worlds" (of what ever type we define)?

The short answer: Time travel changes the rules

The long answer depends on what you mean by "detection" and which time loop resolution mechanism is considered. There are at least two of them. I'm ignoring any problems they have here, for brevity. The short long answer for both is that they introduce nonlinearities into the time evolution. The long long answers:

In Lloyd's model, which is the one you get when you simply apply the usual rules of Quantum Field Theory on a fixed spacetime with loops in it, you detect them only as you usually detect alternate worlds, by their interference patterns that influence the final outcome. The difference that the possible time travel makes is that even those alternatives that usually would not cause interference because they're too far apart can now interfere. The trace operation can mash them together, and it ensures global consistency. An external observer will always see the same thing coming out of the time machine than later gets in. Still, in this model, unless you actively set up paradoxes in some worlds, interference between them is as weak as ever, at least after averaging over macroscopically indistinguishable initial states.

In Deutsch's model, it's simpler. You actually travel back into a different world than you started from. What an external observer sees leaving the time machine does not need to match what he later sees entering it in the future. There are different ways to interpret the equations, though, and the alternate worlds are not necessarily the ones from MWI. You can also see them as actual clones. Of course, the interpretation does not matter for the end result.

Needless to say, if we get hold of a couple of closed timelike curves, we can experimentally distinguish between the two alternatives, they are not equivalent. In the context they are formulated in, there is nothing that makes one preferable over the other, though I imagine Lloyd's trace method would mend easier with whatever Quantum Gravity theory we come up with.

PS, Copenhagen cannot define a "measurement" and MWI cannot define a "split". It's the age old problem of two equal, but incomplete, solutions to a problem. Why argue over them?

Nobody is arguing over them here, you just keep asserting they are both incomplete based on falsehoods Again. We now quite clearly know what a measurement is, at least we can construct an operation for every quantity we want measured that definitely is a measurement. If you entangle your system with a measuring apparatus and leave the apparatus alone for the rest of the experiment (no interaction with the system or other measurements, no quantum erasing, it helps if it is a macroscopic apparatus, but not a requirement), you measured it. Likewise, a split is just writing the state as the sum of two or more states. Every measurement in the above sense comes with a well defined, useful split you can do.

Yes, for a given setup, it's not always easy to tell whether it is to be considered a measurement and if yes, of what. That's also true in classical physics and general statistics, so I would not hold it against QM.

[Technical Ben]

OK, that's interesting. Though it is many "what ifs" far from observation. It requires CTC, fixes to the "collapse" or "superposition" (lack of objects) problems in MWI before being remotely possible. I'm not sure if it brings up new problems, like energy conservation or other problems (Hawkins suggests virtual particle build up would prevent any travel due to their pressure against the time traveller, etc).

My own understanding would be it gets rather blurry in how we are "time travelling" here. If we create a new branch of the universe, but alternate to ours, how is it different than "last Tuesdayism"? Or in this instance creating the universe to look like last Tuesday 20,000 bc?

Or in time travelling to alternate universes, how are we not detecting these MW branches and breaking the "never detect them" rule?

All this would be more understandable if it was Multi System Interpretation or similar. As branching off entire "universes", that is branching beyond all possible detection and all matter/energy in existence repeating, is rather not "simple" from this perspective.

Edited November 27, 2014 by Technical Ben

[Z-Man]

OK, that's interesting. Though it is many "what ifs" far from observation. It requires CTC, fixes to the "collapse" or "superposition" (lack of objects) problems in MWI before being remotely possible. I'm not sure if it brings up new problems, like energy conservation or other problems (Hawkins suggests virtual particle build up would prevent any travel due to their pressure against the time traveller, etc).

Or in time travelling to alternate universes, how are we not detecting these MW branches and breaking the "never detect them" rule?

Well, yeah, we haven't observed anything close to it, so all we can do right now is explore the mathematical possibilities. No fixes are required, actually: What you think are "rules" for MWI are actually consequences of the way time evolution usually works. You take an in-state and via a linear and unitary operation, you transform it to an out-state. A lot of the rules of QM we take for granted derive from that: the no-cloning theorem, the no-communication-via-entagled-states theorem and conservation of total probability. And locally, time evolution is still unitary. However, globally, it seems like it goes down the drain, just like causality itself: for the larger region of spacetime under consideration, it is lost in both models. And without unitarity, the consequences also get invalidated.

And yes, we are ignoring those potential problems. They may be avoidable. Energy conservation is not a problem in the given context: it's another thing that is valid locally, but does not have to be true globally.

My own understanding would be it gets rather blurry in how we are "time travelling" here. If we create a new branch of the universe, but alternate to ours, how is it different than "last Tuesdayism"? Or in this instance creating the universe to look like last Tuesday 20,000 bc?

That's invalid interpretation of calculation steps as reality. Yeah, Deutsch's model looks a bit like that: "First, you take the state of the universe. You use that continuously to feed the time loop until it has settled down. Then, you take the state of the loop and set it as the initial state of the loop in the real universe, then you run that." But that's not relevant. You could just as well interpret the regular time evolution as "Next Tusedayism", because it allows you to progress time one week at a time, each week copying the end state of last week to the initial state of next week. What matters are the results that come out, and the proper way to analyze them is to apply the formalisms (which are interpretation free and don't care whether you use MWI or Copenhagen) to test systems. And then you interpret the result like you would do with any other out-state. You have to treat the region with the time loop as a black box. And to determine how events inside the loop influence the outcome, you play with different in-states or time evolutions and see how things change.

The time travel aspect simply comes in because we partially feed the out-state back into the in-state; it's similar to a regular feedback loop.

All this would be more understandable if it was Multi System Interpretation or similar. As branching off entire "universes", that is branching beyond all possible detection and all matter/energy in existence repeating, is rather not "simple" from this perspective.

It is actually that, the multiple worlds/alternate universes term is just... marketing. Only very rarely is there a need to consider the whole universe. Usually, the states of the photons on your experiment table are all you need to model in your system. Time travel considerations get messy there because you need to include everything that can potentially loop. Edited November 28, 2014 by Z-Man

[monstah]

700kg could be considered vast, based on the fact we have 0g of the stuff

This sort of sentence always cracks me a smile

[Technical Ben]

Thanks Z-man. If MWI is just about systems, and no one system is necessarily "all the universe" it makes much more sense. I can easily have 2 of me (as said, philosophical questions aside), we call that a "twin". Creating 2 objects is trivial, suggesting the creation of everything after every instance across every possibility, is rather the oppoisite.

AFAIK Last Tuesdaism is still a problem we cannot pat under the rug. Do we not still need a time like dimension? If I have no time like dimension, then how can I say I am time travelling? This also seems similar to the "no objects/collapse" problem. As if we have no collapse, we have no "moment" and we have no dimension of time. We instead have either a totally random universe, or a totally stationary universe.

Would it be that in Copenhagen, we say "we observe the collapse, thus we calculate it via event a to b to c". It has no problem with what form reality may take. It just comments on the observation.

However, MWI says "all events happen, Thus we calculate it via event abc and bca and cab". It states all branches contain all events all the "time". A branch where the universe started 1 second after ours is indistinguishable, we then also have the branch where the universe is starting today!

So in MWI nothing "happens", as we cannot enforce a definition of an object or an event. We can never make an "observation" of our universe, let alone any other branch in the MWI model. While Copenhagen has other problems, it does side step this question by just saying "it is what ever we (or the non anthropomorphic "particle/detector/apparatus") observe".

I'm not sure if that is similar to the Bourne probabilities requirement. As in MWI they are assumed, but both not shown emergent in the calculations and in theory can never even be detected in a MWI universe!

But again, Copenhagen requires only we observe them, it is not required to derive them... yet.

[Z-Man]

What do you mean by 'timelike dimension"? In the context of QTF on a fixed background manifold, we have that as the interior of the future lightcone at every event (spacetime point). What we no longer have since 1905 is a meaningful global time coordinate. Usually, one can still at least define a global coordinate that is timelike in the sense that it is smooth and always increases along every futurebound timelike or null world line. But even that goes out of the window if there are closed timelike curves. Ultimately, you'll have to settle for the definition that "time is what the clock measures". That's why we consider the Planck time to be the smallest possible time: Any clock that could measure shorter times would need to be so massive/energetic and small that it would collapse into a black hole. And it also leads to a good definition of time travel: A situation where a non-clonable system containing a clock can be influenced by itself with a more advanced reading on the clock. And the two prescriptions for resolving time loops fit. For a suitable definition of 'itself' in the case of Deutsch's method.

That a model appears static and not moving is not a new problem, either. A pendulum is described as having the motion x(t) = A sin(w t) + B cos(w t). That's one static function (with parameters). No actual movement visible, nothing is happening. Is there anything specifically new here?

And I'm afraid you can't apply the Copenhagen measurement prescription inside a time loop (or, for that matter, a non-reversible MW split). It is not at all clear how it would be supposed to work; it applies to an external observer doing the measuring, and who would be external to the time loop, yet close enough to measure something caught in it? Instead, if you want a measurement from inside the loop, you have to do the complicated thing and model the measuring apparatus as part of the system. It can be a mock apparatus that could not exist in reality, it just needs to be able to measure and store the same thing. Then, you model how the measuring apparatus goes out of the time loop. Only after all is over, you measure the state of the measurement apparatus with your chosen interpretation of QM. And it can matter a whole lot more than normal what you measure and how you treat your equipment. In Deutsch"s method, it matters whether you push your apparatus through the time loop or not after it has done its measurement. In Lloyd's method, what you measure has a lot more influence on the future of the system than usual, and quantum erasers no longer work.

[Technical Ben]

I'm still not convinced. If something requires an addition that we have not yet observed, it's helpful in telling us how to find it, but not helpful until it's found. So without CTCs being observed, we are in the same boat prior to observing a Higgs Like particle, or even less supported "grand unification theories". They can look good on paper, but we need observations, confidence ratings, predictions (with results) and secondary or even tertiary verification.

I saw some nice honesty with the data and facts when it came to the Higgs theory, and even the "detection". With many saying "we found something, Higgs like, and we are learning more".

If we need CTCs, negative energy and infinite energy in the most far out (as in distance from our current observation, not as in "wrong") theories for FTL, then I doubt it will come to anything.

Hopefully we will see a better way to integrate the mechanics in the future. When we had certain predictions fit perfectly for the moon and the stars, but did not relate to both, it took newtonian like mechanics to understand it. When we had perfect predictions for the motion of planets, but then saw Mercury out of "sync" it took relativity to understand it. Perhaps there is one more aspect in QM like observations to make it match up with other systems we observe?

[lazaros]

On 11/18/2014 at 5:20 AM, K^2 said:

Causality is only locally relevant in General Relativity. FTL is not a violation by any measure. Otherwise, Universe wouldn't be expanding at FTL speeds.

The biggest inaccuracy of KSP's Alcubierre Drive is that it doesn't properly take into account curvature due to gravity. You drop out of warp traveling at the same speed that you went into the warp at, but the frame of reference is accelerated. To account for this properly, author should have carried out parallel transport of the velocity vector in a curved space-time. Naturally, since gravity in KSP is a patched-conics approximation, it would be entirely acceptable to use Schwarzschild Metric for parallel transport as well, patching at SOI boundaries as necessary.

There are a few minor issues as well. Energy consumption, visual effect of the warp, etc. But these could be chalked up to sacrifices in the name of gameplay and limitations of the rendering engine.

The "local" relevance is important since traveling using a ship is... a local event for the ship+cargo. The expansion of the universe used as proof of FTL is a non sensical argument: the time space continuum can have an expanding boundary or it might not. Beyond that hypothetical boundary there might be space/time or not. Whatever you believe or think you know, whichever ideas you adhere carry a set of consequences. Causality is not a physics or engineering problem: it's theology. At some point a new type of physical laws might give us a way to achieve FTL. For now we need to use ideas based on the factual reality we comprehend. 5℅ 10℅ c is enough to reach our closest stars in a few decades. It can also make expanding humans to our solar system viable, efficient, and unfortunately profitable. 

I studied physics, got my Msc, my PhD because I wanted a lightsaber and my own Millennium Falcon. I can still dream and hope. 

 

On 11/18/2014 at 5:20 AM, K^2 said:

Causality is only locally relevant in General Relativity. FTL is not a violation by any measure. Otherwise, Universe wouldn't be expanding at FTL speeds.

The biggest inaccuracy of KSP's Alcubierre Drive is that it doesn't properly take into account curvature due to gravity. You drop out of warp traveling at the same speed that you went into the warp at, but the frame of reference is accelerated. To account for this properly, author should have carried out parallel transport of the velocity vector in a curved space-time. Naturally, since gravity in KSP is a patched-conics approximation, it would be entirely acceptable to use Schwarzschild Metric for parallel transport as well, patching at SOI boundaries as necessary.

There are a few minor issues as well. Energy consumption, visual effect of the warp, etc. But these could be chalked up to sacrifices in the name of gameplay and limitations of the rendering engine.

 

[K^2]

On 2/25/2024 at 11:57 AM, lazaros said:

The expansion of the universe used as proof of FTL is a non sensical argument: the time space continuum can have an expanding boundary or it might not. Beyond that hypothetical boundary there might be space/time or not.

There are entire galaxies, observed and measured to be receding from us at the speeds exceeding the speed of light. It's not about the boundary. Stuff in the universe is moving relative to other stuff in this universe at superluminal speeds. This isn't some hypothetical on a napkin. It's a firmly established cosmological fact.

On 2/25/2024 at 11:57 AM, lazaros said:

Causality is not a physics or engineering problem: it's theology.

Causality is a mathematical statement. There are concrete theorems and several notable conjectures that are yet to be proven within a framework. If you don't understand it, all it signals is limits of your education on the subject.

On 2/25/2024 at 11:57 AM, lazaros said:

I studied physics, got my Msc, my PhD because I wanted a lightsaber and my own Millennium Falcon. I can still dream and hope. 

Cool. I was doing research in particle physics. That is, I specifically worked with the concept of matter propagating at energies where space-time metric is the ruling factor, and understanding time ordering is crucial in getting correct results that match experimental data.

You can dream all you want. The measurement precision on QM and GR set the expected scale limitations on where these break down. Classical mechanics was breaking down at sizes much larger than an atom and masses smaller than these of a planet. We could work at these scales and exploit these violations. The QM holds for many orders of magnitude below the scale of any known particle, and at masses exceeding these of galaxies. Humanity isn't going to reach these numbers. Even if there is a higher order theory that is more correct, we aren't going to see the difference, because it'd take several times the energy of the known universe to do anything that does. So unless you want to believe in magic fairies that break down the space-time barrier for us as some sort of a favor, we are going to have to work within the confines of the theory that we have at least until we make crossing the universe as easy as sending a GPS satellite to orbit. Because that's a prerequisite for getting to these scales. The best we can hope for is understanding the theory we have a lot better and make full use of it.

We aren't tapping into a fraction of the near-magical bull crap that QM and GR actually say are possible. Which already covers FTL, wormholes, teleportation, time travel, computational capacity that borders on omniscience, and more. It just has to follow the rules that have already been firmly established. You can't flap your arms and fly despite the fact that we've figured out jetpacks. No matter how much you want to imagine it.

[darthgently]

43 minutes ago, K^2 said:

There are entire galaxies, observed and measured to be receding from us at the speeds exceeding the speed of light. It's not about the boundary. Stuff in the universe is moving relative to other stuff in this universe at superluminal speeds. This isn't some hypothetical on a napkin. It's a firmly established cosmological fact.

But apparent velocity via the expansion of space time is not the same frame as relative velocity.  Neither our galaxy nor the distant one are exceeding the speed of light in their patch of space time. 

Trying to build a space craft drive based on changing space time seems to me like instead of designing a jeep one designs a lever that will rotate the planet underneath one so your desired location arrives underneath you.  Impossibly big lever and untold unintended consequences if you use it and far easier to drive a jeep, fly a jet, or even walk and swim.   Much easier on every other living thing on the planet also

Some human journeys are much bigger than the individual, and their lifespan, or thousands of lifespans, and that is the way it is.   Just our journey of scientific discovery is far bigger than a single human lifespan of exploration, but we do it anyway.  Generation ships will be the only way open to us I predict

On 2/25/2024 at 2:57 PM, lazaros said:

There are a few minor issues as well. Energy consumption,

Lol.  The required energy budget is not what I'd call minor.  Planet consuming energy levels would be required and the magical ability to contain and engineer that energy.   This is why I'm a hard, near future, sci-fi fan.  Anything else just seems silly and whimsical after all the decades I've been enjoying the genre at this point.  Where's the beef?  Ha

[Flavio hc16]

Holy excrements guys...you really relived a 10 years old post? Like seriously?

[darthgently]

13 minutes ago, Flavio hc16 said:

Holy excrements guys...you really relived a 10 years old post? Like seriously?

I just look for recent activity and replied accordingly.  But maybe time dilation is involved for K^2 and not his fault.  Alcubierre drives have side effects like this I gather

[K^2]

10 hours ago, darthgently said:

But apparent velocity via the expansion of space time is not the same frame as relative velocity.  Neither our galaxy nor the distant one are exceeding the speed of light in their patch of space time. 

How is that different for motion between the Earth and the Moon? Or between a starship and a star it's traveling to? How do you measure relative velocities across empty space? You can't put a little turbine out and measure how fast the "space" moves past you. That's the most fundamental principle of relativity.

Mathematically, you fill the space in between with probe particles that are infinitesimally close to each other, and adding up all the relative velocities across the chain. (Because remember, motion is relative so if the chain is "static" relative to you, it might be moving relative to me.) If you do that between distant galaxies, you can't tell the difference between the two galaxies moving apart or the space in between expanding.

In practice, you send a beam of light instead. Or rather, you just wait for it to happen naturally, because it happened billions of light years away and billions of years ago... But regardless, with light, you again run into the same problem. Whether the object far away is receding or the space in between is expanding, you'll get the same result. Motion due to expansion is true relative motion, and it is true relative motion at the superluminal speeds for the galaxies.

10 hours ago, darthgently said:

Trying to build a space craft drive based on changing space time seems to me like instead of designing a jeep one designs a lever that will rotate the planet underneath one so your desired location arrives underneath you.  Impossibly big lever and untold unintended consequences if you use it and far easier to drive a jeep, fly a jet, or even walk and swim

What if I told you that if you write it in the framework of GR, properly accounting for frames of reference, effects of driving a vehicle across the planet, and saying that the vehicle's wheels cause the planet to spin underneath with vehicle staying put are identical?

It's just a mathematical perspective. We happen to use math that explains this type of motion through space-time geometry, because that's the easiest thing for us to put into formulas. There is an equivalent formulation of it under which you just "drive the Jeep". That is, fly the ship at FTL speeds. It's just the kind of math that makes the epicycles of a Geocentric model look quite reasonable in comparison.

 

There are other problems with the Alcubierre Drive, of course. I actually missed the mark a bit on saying the behavior has to be different in curved sapce-time. I had time to revise my understanding in the years since... It's a lot worse. The Alcubierre Drive just straight up doesn't work if it carries any non-zero mass inside of it. Even in flat space-time. The negative energy of the bubble walls has to cancel the mass energy of the ship, or the bubble starts radiating gravity waves. So how do you avoid angular momentum conservation problems when warping around a star system? Super easy, barely an inconvenience. You make sure that the total mass that gets transported is zero. Where you get that negative mass at origin and what you do with it at the destination is left as an exercise to the reader. 

7 hours ago, darthgently said:

But maybe time dilation is involved for K^2 and not his fault.  Alcubierre drives have side effects like this I gather

*Her. I refuse to corroborate on whether this has anything to do with the effects of FTL or time travel.