In the last post, I continued a discussion of time travel and pointed out that any time travel into the past would result in facts/events that become permanently embedded in the information structure of the universe. I’ve had several more insights since then about the (im)possibility of time travel. Here they are in random order:
Traveling into the Future
The examples I gave were all about changing the “rate” of travel into the future, such as with drugs and relativistic time dilation. I should be clear that it is impossible to “skip” to some time in the future, such as with Doc Brown’s Delorean time machine. You can’t instantaneously skip to the future because determinism is false, so you have to just wait it out for random/quantum events to actually occur. In other words, there is no way for the universe to know what the year 3000 looks like until it actually goes through the (time-consuming! haha) effort of evolving through that time. Lots and lots of new information has to be created through, for example, quantum events, and the universe doesn’t “know” what quantum possibilities exist in the year 3000 until the events of the year 2999 have happened, and the quantum possibilities in the year 2999 depend on the events of the year 2998, and so forth. Of course, you can consciously skip over time (by sleeping, etc.), but there will still be evidence in the universe of time passage, including in your own body.
Traveling into the Past
The explanations/arguments I have in my last post about why time travel to the past is impossible relate to how changes will affect the present (or the future of the past), which is problematic for reasons I’ll explain more momentarily. But someone might say, “I don’t need to actually travel to the past if that’s impossible. A simulation would be fine.” Obviously, simulations are possible to some degree. Setting aside that a realistic simulation would itself require enormous computational resources, any simulation would be limited to a tiny selection of information. There is no way to recreate or accurately simulate the past because the vast majority of the information in the universe (correlations/entanglements) is fundamentally inaccessible. No matter how many experiments you do, the vast majority of the information in the universe cannot, even in principle, be known. (The reason for this is related to quantum no-cloning.)
I made the mistake in the last post of confusing the analysis with questions about free will. After all, imagine I don’t have free will and instead whether I travel into the past in my time machine depends on some quantum event 10 minutes from now. The same problems arise because, at this moment, there is no fact about whether or not I’ll travel into the past, which means that whether or not I traveled into the past cannot already be embedded in the information structure of the universe.
The word “change,” particularly when philosophers and physicists discuss time travel, is misused, abused, and confused. The word “change” itself is always relative to a particular time. “Yesterday, I changed my mind.” “Last week, I changed my haircut.” “At time t0, the particle’s trajectory changed.” But people who discuss time travel often say things like, “Tomorrow, I will change what happened last year,” or “At time t2, what happened at time t1 was changed.” This is nonsense. When was the event changed? At t1 or t2?? Use of the word “change” in this fashion is inherently contradictory.
But here is a much bigger problem with the notion of time travel into the past: it assumes that the universe evolves deterministically. We tend to think that if we were to change the past just a little bit, then the present would also change only a little bit. Or maybe it would change a lot (thanks to chaos) but we assume that the changes to the present depend only on the changes to the past. Similarly, if we could manage not to change the past at all, then the present would remain exactly the same. This is a natural assumption – in fact, it was one that I hadn’t even questioned until today – but it is wrong.
That’s because the information content of the universe is increasing over time, thanks to the indeterminism caused by random/quantum events. So if we were able to travel to the past, then the past we would experience would have less information and there is no guarantee that it would evolve into the present that we actually currently experience. In fact, it would almost certainly NOT evolve into the actual present. And if you changed something about the past – anything at all – then different quantum information must be created because different quantum possibilities would arise.
Here's a simplified example showing the problem, in which I am somehow able to travel back to time t0. Assume:
· A speck of dust went in direction A at time t0 but if I time travel to t0 then it will/would/did go in direction B.
· In a nearby quantum experiment, an object in quantum superposition state x|X> + y|Y> could be detected at location X or Y (where x and y are complex amplitudes). Then, at time t1, it was actually detected at location X.
Problems & Questions:
· Is there any good reason to think that, if I time travel to t0, the object will still be detected at X at t1?
o No, because if there was, then that means the object was already in state |X>, not a superposition.
o Similarly, the outcomes of every possible quantum event in the universe could be different. There is essentially zero chance that the universe I currently experience, or anything like it, will evolve if the embedded information of events since time t0 is erased.
· If I am right about universal entanglement (i.e., the speck and object are already well correlated), wouldn’t the changing of direction A to B also affect the amplitudes x and y, even if only slightly?
o If so, then there certainly would be no reason to think that the object would be detected at X at t1. Further, correlation information with objects and fields throughout the universe would instantaneously (nonlocally) change. Event outcomes would change, new possibilities would arise while others would be foreclosed, and so on.
· What does locality have to do with this? Locality doesn’t matter regarding universal entanglement, because quantum correlations are nonlocal. But even setting that aside, if the speck goes in direction B instead of A, then new (quantum) possibilities for creations of photons are created, and those new photons could affect the experiment, which clearly would affect the amplitudes x and y.
 This is more linguistic evidence of the contradiction inherent in backward time travel. If I time travel to time t0, do I say that the speck of dust will go in direction B? Would go? Did go? WHEN did I change its direction? Now or at time t0?