Can retrocausality solve the puzzle of action at a distance?

I consider this view supported by relativity where, as you mention, technically a photon is an atomic event from its own point of view. There isn't any before or after in its own frame of reference.

What we call time would then be a secondary phenomenon on top of this much more complicated (from our point of view) time organization of the universe. (However, I reject the idea that that makes "our" time an illusion, or some BS like that. Our human conception of time is perfectly real. It just may not be fundamental. But even if we discover that "real time" is something like what I'm hypothesizing here, human time will still be as real as ever.)

What is not clear to me is whether or not this could be set up in a reasonable manner. It's on my long list of things I'd love to have time to try my hand at. It's feasible, you just need to start with maybe a 5 or 6 qubit system that interacts in this "real time" manner in the counterintuitive "retrocausal" manner, but for which there is a point of view in which it all happens forward in this "real time" metric. You don't have to simulate an entire universe to show this is a feasible idea, if you can establish it at that scale, it is merely a matter of extreme complication to assume the universe could work that way.

There's also an interesting implication for the "deistic" point of view. If this is how quantum mechanics "really works" then it becomes relatively easy to imagine that the Great Simulator is simulating the universe, but actually has little to no power to influence it, because from his point of view during the computation the entire universe is a tangled mess of everything essentially happening at once (from our point of view). The only thing the Great Simulator might be able to do is examine the final state of the universe after it has finished computing, at which point the entire history of the universe would be open to him, but to affect the state of the universe would be wildly more computationally expensive than the simulation itself, because any attempt to insert or modify even a single qubit's worth of change would essentially require the entire thing to be recalculated again as the changes propagate in (from our point of view) all the way forward and backwards in time again. Depending on the nature of "true time", though, it might be possible, depending on exactly how it ends up working.

From the perspective of the photon (really kind of silly since the photon can't have any mass, and therefore no observer), there is no time or distance traveled. It takes 0 time to travel 0 distance.

This is not correct. The correct statement is that, in SR, the photon's worldline is a null worldline. But the concept of "proper time" is not well defined on a null worldline, so you can't say that "no time passes". In fact, in any inertial frame, the time of the detection event will be later than the time of the emission event, so time certainly does pass between the two events in the only sense in which "time between events" can be consistently defined in SR.

Also, the photon's state does not have to be constant along a null worldline.

>is this what the Paris Zigzag does - treat the state of the photon and it's twin as a single state

No. The "Paris Zigzag" just says that causality can flow "backwards" along null worldlines, from what we call the "detection" event to what we call the "emission" event, instead of just the other way around. All of what I said above still applies.

For a handful of others in this thread:

While it's strictly true that proper time is hard to define usefully for a classical massless particle, you can still parameterize a null curve any way you want, depending on what it is you're trying to do.

In flat spacetime you're rarely trying to eke out a transformation much more complex than a Lorentz transform (the matrix for which involves \gamma, which is undefined for v=c), and so you're stuck with ds = 0. On the other hand, your objects on lightlike curves are probably light itself (classical or otherwise), so you'd generally look to something like an affine parameterization like k^b\Delta_{b}k^a=0 and k_{a}k^a=0 where k^a(x) are the components of the photon's momentum, or (aiming for E=pc) parameterize on the wave vector k (where v=c implies k^2={\omega}^2, where \omega is the wave angular frequency).

Other sensible choices of affine parameter are available, and typically give you freedom to pick out a momentum vector k^\mu = \dot A^\mu where A is the affine time; usually you'd do this so as to match the momentum-energy at the emitter or absorber (or the redshift along the null curve).

Basically, hidden variables models can't explain the results of some of our experiments and no-one's found a way to make the two match up.

Reference: https://en.wikipedia.org/wiki/Bell's_theorem

It is not as simple as you'll get with macroscopic things, but yes it is not as crazy as some people imply either.

---

...The hypothesis behind invisible writings was laughably complicated. All books are tenuously connected through L-space and, therefore, the content of any book ever written or yet to be written may, in the right circumstances, be deduced from a sufficiently close study of books already in existence. Future books exist in potentia, as it were, in the same way that a sufficiently detailed study of a handful of primal ooze will eventually hint at the future existence of prawn crackers.

But the primitive techniques used hitherto, based on ancient spells like Weezencake's Unreliable Algorithm, had meant that it took years to put together even the ghost of a page for an unwritten book.

It was Ponder's particular genius that he had found a way around this by considering the phrase, "How do you know it's not possible until you've tried?" And experiments with Hex, the University's thinking engine, had found that, indeed, many things are not impossible until they have been tried.

Like a busy government which only passes expensive laws prohibiting some new and interesting thing when people have actually found a way of doing it, the universe relied a great deal on things not being tried at all.

When something is tried, Ponder found, it often does turn out to be impossible very quickly, but it takes a little while for this really to be the case–in effect, for the overworked laws of causality to hurry to the scene and pretend it has been impossible all along. Using Hex to remake the attempt in minutely different ways at very high speed had resulted in a high success rate, and he was now assembling whole paragraphs in a matter of hours.

"It's like a conjurin' trick, then," Ridcully had said. "You're pullin' the tablecloth away before all the crockery has time to remember to fall over."

And Ponder had winced and said, "Yes, exactly like that, Archchancellor. Well done."

And that had led to all the trouble with How to Dynamically Manage People for Dynamic Results in a Caring Empowering Way in Quite a Short Time Dynamically...

- Terry Pratchett, The Last Continent

But I'd understand why you had to write that anyway :-P