What would a time machine look like?

[Jaxodius]

Hello

Since we know that time travel is possible, what would the time machines look like? How large would it be? For the moment, I am not concerned with how it would be powered to bring it up to a speed fast enough , but construction wise what would it look like? will be like a cylinderical tube? or a small room? or a space just small enough to fit a human?

 

[whybother]

Hello

Since we know that time travel is possible

I'm going to stop you right there...

 

[Jaxodius]

Am i wrong? I thought if you travel at the speed of light in a circle and come to a halt where the circle started, you would have moved forward in time. Is this wrong?

 

[whybother]

Am i wrong? I thought if you travel at the speed of light in a circle and come to a halt where the circle started, you would have moved forward in time. Is this wrong?

Yes, actually. Massive objects can never be accelerated to the speed of light, for starters (see the many discussions here and elsewhere on special relativity for me on that). But also rather significantly, I assume, when you say "in a circle" you at least mean a closed loop in spacetime. Timelike loops are forbidden by causality in special and general relativity, meaning no time travel.

If you are interested in models of cosmology that allow for 'time travel' (and still, being good physics, preserve causality), then look into Godel universes or other topologically nontrivial spacetimes. This still does not allow for 'time machines' though, just a structure of spacetime that could, in theory, allow one travel 'backwards' in time. None of these models are actually viable for a universe with matter, really... and almost all relativist throw out any spacetime that allows for those closed timelike loops like you want, for being unphysical (and very unlike the universe we perceive).

 

[A.T.]

Am i wrong? I thought if you travel at the speed of light in a circle and come to a halt where the circle started, you would have moved forward in time.

I'm moving forward in time right now, using this device:

 

[whybother]

I'm moving forward in time right now, using this device:

Hahaha, I read the original post as moving backwards in time. Yes, a time machine to move forward in time can look... pretty much like whatever you want lol

 

[Jaxodius]

I was wondering how i could have got it so wrong. I should have mentioned this in the first post. Yes i was talking about a time machine that travels into the future.

Ok so as long as the machine can be accelerated fast enough, it could look like a room, a house, a car, a coffin or any number of enclosed spaces?

 

[whybother]

I was wondering how i could have got it so wrong. I should have mentioned this in the first post. Yes i was talking about a time machine that travels into the future.

Ok so as long as the machine can be accelerated fast enough, it could look like a room, a house, a car, a coffin or any number of enclosed spaces?

No, no, you are still very incorrect if you mean 'travel into the future faster than you would be doing otherwise'.

 

[Jaxodius]

Oh, so no matter how fast you go, you would never travel into the future any faster than you would just sitting in your computer chair?

 

[A.T.]

Ok so as long as the machine can be accelerated fast enough, it could look like a room, a house, a car, a coffin or any number of enclosed spaces?

It doesn't need to be enclosed, if you don't mind the vacuum in space. If it accelerates you away and then back to the start point, you will have experienced less time, than the start point. So you would reach the future of the start point faster than waiting there.

 

[Dmitry67]

Timelike loops are forbidden by causality in special and general relativity, meaning no time travel

1. Causality is a consequence of physical laws (in the regions of relatively flat spacetime), but not an axiom
2. AFAIK, closed timeloops do exist around kerr singularities
3. Causality can be preserved in the timelike loops (Novikov's principle)

 

[DaveC426913]

Woah there fellas... you're barking up the wrong metaphor.

What Jax is referring to is a relativistic shuttle.

Head out to Sirius at .999c, turn around and come back at .999c and Earth will have aged significantly while you will have only aged a few years.

He is correct that, using a spaceship to travel at near the speed of light in a loop, you would effectively travel into Earth's future.


Jax, you didn't have it wrong, it was just interpreted the wrong way - too many people thinking you were asking something other than what you were.


And the answer is: it would look like a spaceship capable of sustaining its occupants for a few years or decades, with an engine capable of accelerating continually up to relativistic speeds, and carrying a nigh-unlimited supply of fuel.


(Joe Haldeman used this in The Forever War, to have the heroine "wait" for centuries (Earth-time) while the hero "caught up with her").

 

[stevebd1]

According to GR, there is massive time dilation around super compact objects such as neutron stars and black holes, if you were to hover near the event horizon of a large static BH where the tidal forces were small, the outside universe would appear to speed up and technically, once you leave the BH's gravitational field, you would have traveled into the future.

But what about the ergoregion of a rotating black hole? Space is being dragged in the azimuth plane within this region faster than the speed of light and proper time according to Kerr metric which becomes zero at the static limit (the outer edge of the ergoregion) becomes negative between the static limit and event horizon, implying that maybe, if you entered the ergoregion and didn't fall into the black hole, you would travel faster than the speed of light relative to the outside universe, which, according to SR, seems to imply a different sort of time travel (i.e. into the past).

 

[Jaxodius]

Head out to Sirius at .999c, turn around and come back at .999c and Earth will have aged significantly while you will have only aged a few years.

Thank you Dave. That is exactly what i was trying to say.

with an engine capable of accelerating continually up to relativistic speeds

That is my thought exactly. I really believe it is simply a matter of time before such a machine is built. If 2 assumptions are made, and they are not far fetched by any means, then it is absolutely possible to build a machine that will take ppl into the future. The 2 assumptions are:

1] we can use a source of fuel that will be there for a very very very very long time ( the sun? )
2] we build an engine with no wear and tear. ( future ion drives? )

if you entered the ergoregion and didn't fall into the black hole, for what ever period of time you remained in there, you would travel faster than the speed of light relative to the outside universe, which, according to SR, seems to imply a different sort of time travel (i.e. into the past)

I have been thinking about that exact same thing. But as far as our current understanding of light is, we cannot travel exceed the speed of light. But still, we know how (relatively) easy it is to travel into the future ( oh man imagine saying this 50 years ago... ). There HAS to be a way that is just as (relatively) easy to travel into the past. There JUST HAS TO BE.

 

[alxm]

Wouldn't someone have already traveled back in time to show us what one looks like?

My guess would be: "Surprisingly similar to a DeLorean".

 

[JesseM]

But what about the ergoregion of a rotating black hole? Space is being dragged in the azimuth plane within this region faster than the speed of light and proper time according to Kerr metric which becomes zero at the static limit (the outer edge of the ergoregion) becomes negative between the static limit and event horizon, implying that maybe, if you entered the ergoregion and didn't fall into the black hole, you would travel faster than the speed of light relative to the outside universe, which, according to SR, seems to imply a different sort of time travel (i.e. into the past).

GR always reduced locally to SR for a freefalling observer (the equivalence principle), so nothing can ever travel faster than the speed of light in a local sense unless tachyons exist. The discussion of the ergosphere here suggests that when they say space is being dragged faster than the speed of light, they mean that even something moving at the speed of light in the direction opposite to the direction of rotation cannot mantain a fixed angle in whatever global coordinate system is normally used to describe rotating black holes. But I know physicists have analyzed the problem of where closed timelike curves (indicating backwards time travel) might occur, they definitely don't occur in the ergosphere although they could occur inside the inner event horizon if the Kerr solution is accurate (but the Kerr solution has questionable plausibility because there's infinite blueshift of infalling light on the boundary of the inner horizon). This is discussed briefly in the paragraph before section 3 (p. 12 of the PDF) of this paper by Kip Thorne:

Another famous vacuum solution of the Einstein equation that has a non-compactly generated chronology horizon is Kerr spacetime. The exterior of Kerr's outer horizon ($$r > r_+$$ in the usual notation) and the region between the outer and inner horizons ($$r_+ > r > r_-$$) are chronal; the inner horizon ($$r = r_-$$) is a chronology horizon; and the region inside there ($$r < r_-$$) is nonchronal. It is conventionally argued that, although the chronal region is likely to occur in our real universe as the exterior and interior of an old rotating black hole, the spacetime near and inside the chronology horizon will be altered by an instability due to infalling, blueshifted perturbations; and this alteration (hopefully) will prevent CTCs from arising [17].

 

[JesseM]

I have been thinking about that exact same thing. But as far as our current understanding of light is, we cannot travel exceed the speed of light. But still, we know how (relatively) easy it is to travel into the future ( oh man imagine saying this 50 years ago... ). There HAS to be a way that is just as (relatively) easy to travel into the past. There JUST HAS TO BE.

In relativity objects don't really "travel" in either direction of time, they just have worldlines which represent their path through spacetime (similarly, if you draw a line or a curve on a piece of paper, it isn't really meaningful to ask if it's 'traveling up the page' or 'traveling down the page'). The idea of "backwards time travel" is represented by the idea of a worldline that actually curves around and intersects itself, like if you went back in time and shook hands with your younger self...this idea is known as a closed timelike curve, and although you can find theoretical examples of spacetimes that contain them, they all have features which make them questionable as spacetimes that could occur in reality (like the infinite energy on the boundary of the inner horizon of a rotating black hole, or the fact that the whole universe has to be rotating in the Godel metric), and there are also some arguments as to why a theory of quantum gravity (which it's thought that general relativity will turn out to be an approximation to) might remove these possibilities for closed timelike curves (for example see this article on how string theory, a popular candidate for a theory of quantum gravity, might eliminate closed timelike curves).

 

[Dmitry67]

Thank you, very interesting article.
But the summary is "we don't know yet" :)

 

[Phrak]

A time machine would look like a double gimbaled gyroscope, spherical in shape, with a blond actress sitting in the middle. The gimbals are giant hoops, and as they spin around they make powerful wushing noises.

 

[matheinste]

A time machine would look like a double gimbaled gyroscope, spherical in shape, with a blond actress sitting in the middle. The gimbals are giant hoops, and as they spin around they make powerful wushing noises.

My machine, probably an earlier, (or later?) mode,l makes a gnihsuw noise.

 

[Dale]

Wouldn't someone have already traveled back in time to show us what one looks like?

AFAIK all of the GR "time machines" can only go back to when the time machine was built, and usually not that far.

My guess would be: "Surprisingly similar to a DeLorean".

:rofl:

 

[Jaxodius]

Wouldn't someone have already traveled back in time to show us what one looks like?

I have thought about that. Here is what i have. Tell me what you guys think.

http://img26.imageshack.us/img26/3028/howtimetravelworks.jpg

Person A travels back in time, let's say to 2000 BC from 2000 AD. He travels from a timeline called "Time line 1". This timeline is what we are living in right now. If a person can travel back in time, he will go into the past of Timeline 1, but everything after that will change and a new timeline is created. The original timeline carries on, and no recorded appearance of Person A is found from the past. Person A has, through his time travel, created a different version of the timeline , Timeline 2. So the reason there were no recorded history of some traveller from the future is because Person A has created a second timeline simply by time travelling.

Ofcourse, according to this, the implication is that once you travel back in time, you can never return back to Timeline 1 by traveling into the future. If Person A chooses to travel into the future, unless he/she has been extremely careful not to alter events too much ( as in , it is ok if the person decides to step on a few insects but not ok if the person decides to teach people about gravity in 2000 BC ), he/she will come to a future which maybe 99.99% similar to the 2000 AD of timeline 1 but it is NOT timeline 1. It is timeline 2.

 

[stevebd1]

Hi JesseM

Thanks for the response. Are you saying that even though objects that enter a region where space is rotating faster than c, the object itself, relative to the outside universe, will never rotate faster than c and that space in the ergoregion 'rushes' past the object? The most influence being placed on the object is that it may approach a tangetial velocity of c (relative to infinity) as it reaches the horizon where the frame dragging rate increases to infinite.

Also, proper radial velocity for a Kerr black hole is expressed-

$$v=\frac{\sqrt{2Mr(r^2+a^2)}}{\rho^2}$$

where

$$\rho^2=r^2+a^2cos^2\theta$$

and $M=Gm/c^2$, $a=J/mc$

which again implies that proper radial velocity increases over c relative to infinity not only outside the event horizon but outside the ergoregion. I have some idea of the difference between radial and 'proper' radial velocity but what exactly is going on here?

Regarding closed time-like curves, the reduced circumference $(\varpi)$ in Kerr metric is expressed-

$$\varpi^2=\frac{\Sigma^2 sin^2\theta}{\rho^2}$$

where

$$\Sigma^2=(r^2+a^2)^2-a^2\Delta sin^2\theta$$

and $\Delta= r^{2}+a^{2}-2Mr$

which doesn't deviate too much from the coordinate radius (r) until within the Cauchy horizon where it begins to diverge from r significantly and increase exponentially in close proximity of the ring singularity (where r=0 at the edge of the ring). This would give the impression that the reduced circumference would 'wrap round' the ring singularity, overlapping the original position. A simplistic view maybe but it does seem to suggest CTC's.

 

[JesseM]

Hi JesseM

Thanks for the response. Are you saying that even though objects that enter a region where space is rotating faster than c, the object itself, relative to the outside universe, will never rotate faster than c and that space in the ergoregion 'rushes' past the object?

No, when I said things don't go faster than c, I referred specifically to a locally inertial frame--this is a term often used in discussions of the equivalence principle, it basically means you're zooming in on an infinitesimally small region of curved spacetime so the effects of curvature go to zero (at least to the first order). If you're comparing velocity in the ergosphere to velocity outside it then you're not dealing with a single local region, and the coordinate system you're using must be a non-inertial one. Even in flat SR spacetime, it's quite possible for things to travel faster than c in non-inertial coordinate systems, the light speed limit only applies to distance/time measured in the coordinates of an inertial frame.

Regarding closed time-like curves, the reduced circumference $(\varpi)$ in Kerr metric is expressed-

$$\varpi^2=\frac{\Sigma^2 sin^2\theta}{\rho^2}$$

where

$$\Sigma^2=(r^2+a^2)^2-a^2\Delta sin^2\theta$$

which doesn't deviate too much from the coordinate radius (r) until within the Cauchy horizon where it begins to diverge from r significantly and increase exponentially in close proximity of the ring singularity (where r=0 at the edge of the ring). This would give the impression that the reduced circumference would 'wrap round' the ring singularity, overlapping the original position. A simplistic view maybe but it does seem to suggest CTC's.

I don't know how the conclusion of CTCs inside the inner horizon is reached, but I would caution against reading too much into any coordinate-dependent statements like this, there are plenty of examples in GR of things that look really weird in one coordinate system but turn out to just be coordinate artifacts with no physical significance. For example, in the case of a nonrotating black hole spacetime, if you use Schwarzschild coordinates then it takes an infinite amount of coordinate time for anything to reach the horizon so it was once thought there was some sort of singularity there, but later physicists understood that this effect disappeared if you used other coordinate systems on the same spacetime (like Kruskal-Szekeres coordinates), and that it takes only a finite proper time for an infalling observer to reach the horizon, so there is no physical singularity at the horizon. CTCs are a coordinate-independent notion--geodesics that wrap around and intersect themselves--so I don't think you can conclude anything about whether they exist in a given region just by looking at the peculiarities of how a particular choice of coordinate system behaves in that region (and the coordinate system you're talking about may be the most convenient one to use for a rotating black hole, but it isn't the only possible one).

 

[JesseM]

I have thought about that. Here is what i have. Tell me what you guys think.

http://img26.imageshack.us/img26/3028/howtimetravelworks.jpg

Person A travels back in time, let's say to 2000 BC from 2000 AD. He travels from a timeline called "Time line 1". This timeline is what we are living in right now. If a person can travel back in time, he will go into the past of Timeline 1, but everything after that will change and a new timeline is created. The original timeline carries on, and no recorded appearance of Person A is found from the past. Person A has, through his time travel, created a different version of the timeline , Timeline 2. So the reason there were no recorded history of some traveller from the future is because Person A has created a second timeline simply by time travelling.

Ofcourse, according to this, the implication is that once you travel back in time, you can never return back to Timeline 1 by traveling into the future. If Person A chooses to travel into the future, unless he/she has been extremely careful not to alter events too much ( as in , it is ok if the person decides to step on a few insects but not ok if the person decides to teach people about gravity in 2000 BC ), he/she will come to a future which maybe 99.99% similar to the 2000 AD of timeline 1 but it is NOT timeline 1. It is timeline 2.

This is one possible way time travel could work, although more often physicists who speculate about such things imagine that if time travel was possible, there would just be a single self-consistent time travel in which nothing could be "changed", and any influences a time traveler had on the past would have been part of history all along (for example, you might travel back and start the Great Fire of London, in which case it was always true that you were the one who started it even before you went back). The idea that history is constrained to be consistent is sometimes called the "Novikov self-consistency principle", you can read about it here:

http://en.wikipedia.org/wiki/Novikov_self-consistency_principle

It seems to me that while the "branching parallel universes" theory you mention doesn't give rise to any obvious paradoxes, there'd be a lot of ways in which the "rules" governing time travel would probably have to be sort of inelegant and arbitrary. A while ago I wrote up a few paragraphs on my problems with the idea, using the "Terminator" movie as an example (if you don't know the plot, all you need to know here is that in the movie's 2029 there's an evil computer called Skynet at war with humanity, it sends a 'terminator' robot back to 1984 to kill Sarah Connor, the future mother of John Connor, the leader of the human resistance in 2029; in response, John Connor sends back a soldier named Kyle Reese to stop the terminator in 1984).

Whenever I read stories with multiple timelines before, I always assumed the way it worked was this: if you start out in timeline A and you get sent back to some date, say 1066, you'll create a new timeline B that was identical to timeline A right up until the exact moment you appeared in 1066, and which only begins to diverge after that moment. But this assumption doesn't actually work so well in a scenario where you have more than one time traveler sent back to the same time period. If you're starting from a timeline A where there were no time travelers in 1984, and then Skynet sends back the terminator and Kyle Reese goes back after him, then if the terminator arrived a few hours before Kyle, Kyle would be appearing in a timeline that had actually diverged from his own past a few hours before he arrived. This may not seem like a big deal, but if you allow that then you could have crazy situations like Kyle and the terminator going back to 1984 all serious and determined about killing/protecting Sarah Connor, but instead finding themselves in a 1984 ruled by intelligent dinosaurs because some time traveler in 4328 A.D. had decided to take a trip back to the Cretaceous era.

If you stick to the principle that each time traveler finds himself in a timeline that only diverged from his own past after the moment he arrived, then when the terminator and Kyle jumped back from timeline A, I think that'd have to result in two separate timelines: timeline B, where the terminator is the only one to arrive in 1984, he kills Sarah Connor, years later Skynet wins; and timeline C, where Kyle arrives alone and has no one to protect Sarah from. If you take this to extreme, then even if there's a 10^-30 second difference between the moment Kyle's feet appear in 1984 and the moment atoms from his head appear, his atoms would end up scattered between different timelines (and given the relativity of simultaneity in relativity, it's hard to see how to avoid this unless the theory of relativity is wrong).

An alternative to having each time traveler create a new timeline would just be to allow time travelers to appear in versions of the past that had already diverged from their own history before they arrived thanks to other time travelers; in this case, you might imagine taking a God's-eye-view of timeline A from the Big Bang to the end of the universe, taking into account all the time travelers ever sent into the past in timeline A, and then creating a single new timeline B in which all these time travelers appear at the appointed moment, in chronological order. So if the furthest trip into the past anyone made from timeline A was a condemned prisoner who was sentenced to death by being sent back to 30 seconds after the Big Bang, then timeline B would begin to diverge from timeline A 30 seconds after the Big Bang when that prisoner appeared, and all time travelers sent back from timeline A would find themselves in a history that had been influenced by that event.

One other solution to the problem of multiple time travelers heading for the same period would be to have something like a wormhole gate that offers a permanent connection between two timelines as long as it's held open, so if in the 2029 of timeline A you open a gate to 1984 it'll lead to a timeline B which diverges from A at the moment the other end of the gate appears in 1984, and everyone who subsequently steps through the gate in timeline A will find themselves in the same timeline B (and maybe can even return through the gate back to timeline A whose own history hasn't been changed), but then if this gate is shut down in and someone opens a new gate to 1984 in timeline A, this gate will lead to a different timeline C rather than back to timeline B.

 

[DaveC426913]

This is one possible way time travel could work, although more often physicists who speculate about such things imagine that if time travel was possible, there would just be a single self-consistent time travel in which nothing could be "changed", and any influences a time traveler had on the past would have been part of history all along (for example, you might travel back and start the Great Fire of London, in which case it was always true that you were the one who started it even before you went back). The idea that history is constrained to be consistent is sometimes called the "Novikov self-consistency principle", you can read about it here:

The trouble with it is that it precludes free-will. You cannot decide at the last minute to forego the trip back in time to instead have a cup of tea with some cucumber sandwiches under the parasol in the yard.

 

[JesseM]

The trouble with it is that it precludes free-will. You cannot decide at the last minute to forego the trip back in time to instead have a cup of tea with some cucumber sandwiches under the parasol in the yard.

As long as you don't specifically know how the fire started, then you won't feel any problems with free will, if you choose not to visit 1666 that just means you must not have been the one that started the fire. The problem arises when you know in advance what choices you make in your own future. Still, it doesn't seem there's anything fundamentally illogical about this in principle, here's something I wrote on the idea a while ago:

Here's an analogy. Imagine you want to write a computer program to generate a possible chess game. One way is to start with the pieces in their starting configuration, then have the program generate each successive configuration on the next turn from the configuration on the previous turn, using only legal chess moves. But here's another, more elaborate way to do it. have the computer generate an entire series of configurations at once, completely randomly, so it just picks randomly which pieces to put in which positions on which turn. It is very unlikely that the resulting series will look like a legal chess game--a piece might randomly be on a particular square on one turn, but then the next turn randomly be on some totally different square that it shouldn't be able to get to in one move by the rules of chess. But suppose you have access to an idealized computer with nearly infinite speed and memory, and you have it generate a gigantic number of random series this way--if your number is large enough, chances are at least some of the series would just happen to satisfy the rules of a legal chess game. So you could specify that the computer should throw out all series which violate the rules of chess, and be left only with series that represent legal chess games. But since you are dealing with an entire series at once, you could also place other constraints on them, like "throw out all series where white wins", or "show me only series where the black rook checkmates the king in 25 moves", whatever you want. For sufficiently detailed conditions, it might be very hard to generate a chess game that matched them in the traditional way of starting from the beginning and basing each new configuration of pieces on the configuration of the previous turn, but using this brute-force method of generating a near-infinite number of entire histories, and throwing out all but the ones that satisfy your constraints, it's easy to get a game that satisfies any conditions you like without even having to think about it or plan the details of the game.

Similarly, suppose you were using this incredibly powerful computer to generate a simulation of an entire universe--instead of picking some initial conditions and then letting it evolve forward according to some set of laws of physics, you could again specify your "laws" in terms of constraints on entire histories, with the computer generating a huge number of random histories and then throwing out all the ones that don't satisfy the conditions. If the "laws of physics" you pick happen to allow time travel, then obviously any universe that respects the laws of physics locally at every point in spacetime must be globally self-consistent, and the computer will find some histories satisfying this condition. But the computer does not need to have any intelligence to do this, it's just randomly generating a huge number of possibilities until it finds one that satisfies the constraints. From the point of view of a time traveler in this universe, it might seem like the universe was cleverly finding ways to "outsmart" them and thwart their plans, but it would actually be the result of a fairly simple rule, just not a dynamic rule.

In this type of simulation, there might be some random histories in which an older self appears out of nowhere and says something to the younger self, but the younger self never actually goes back in time, so the older self's appearance has no causal explanation and the history is inconsistent; but the computer would just throw away such inconsistent histories. As long as there are some possible histories in which the time traveler's actions are consistent, this imaginary supercomputer will find them and return them as output. We still might ask what a "typical" consistent history would look like--perhaps histories in which simulated time travelers just never get into situations where they learn anything about their own futures would be a lot more probably than histories in which they did learn their future and acted in ways exactly consistent with what they had learned. In fact, in terms of probability, I imagine it would be more likely still that the computer would return histories in which no simulated civilization ever managed to actually create a time machine even if the laws of the simulation allowed for time travel, but this would sort of take the fun out of things.

 

[DaveC426913]

As long as you don't specifically know how the fire started, then you won't feel any problems with free will, if you choose not to visit 1666 that just means you must not have been the one that started the fire. The problem arises when you know in advance what choices you make in your own future. Still, it doesn't seem there's anything fundamentally illogical about this in principle

There is a fundamental flaw. You could easily contrive a situation so that you know what happened (you have all the time in the world to organize and verify it), and then try to not go and do it.

 

[JesseM]

There is a fundamental flaw. You could easily contrive a situation so that you know what happened (you have all the time in the world to organize and verify it), and then try to not go and do it.

But the supercomputer in my thought-experiment would simply throw out any histories where you made that choice, and keep only ones where you didn't know, or where you did know and chose to go through with it (or chose not to but were forced by someone else at gunpoint, or were hypnotized, or someone staged an elaborate conspiracy to make you think you did it using an android lookalike, etc. etc.) This is assuming you believe that your brain is a physical system that would just be one element of a supercomputer simulation of an entire universe at the subatomic level, rather than there being some supernatural element to free will.

 

[Dmitry67]

There is a fundamental flaw. You could easily contrive a situation so that you know what happened (you have all the time in the world to organize and verify it), and then try to not go and do it.

It is easier to think about it in MWI. So we check all possible chains of events, but incosistent sequences are ruled out

The question is - when and how? How should it appear to the observers?

What will actually prevent me from killing my yonger self? Stupid accidents, traffic jams, i will get sleepy, I'll be afraid that if I do something inconsistent then Universe would bluescreens?

And when all these mysterious things would start to happen? The closer to the inconsistent event, the more desperate and mysterious efforts nature has to do to avoid it.