I guess we could just go back and ask Mr. Schroedinger
I have here a version of Schroedinger's box. I don't like killing cats, so I've stuck Hitler inside of it instead.
My box is configured so that sixty seconds after I start the experiment, there will be a 50% chance that Hitler will be dead and a 50% chance that he'll be alive. At that point, I'm going to open the box and broadcast the results to the world.
...
...
...
He's alive! OK, that concludes that experiment.
But wait!, I imagine you saying. I wanted Hitler dead! Don't we all, my friend. Don't we all. But if you assume a model of time travel wherein the past can be changed, how about you just go back in time and so you can observe the experiment again?
So let's hypothetically say that you go back about a minute and a half, to before I've started the experiment. You don't interfere with the process in any direct way: you just quietly wait until it's completed and listen to the results again.
Will the experiment still be random, with a 50% chance of life or death? Does your mere presence in the past have the ability to alter the result, producing a potentially different answer this time around, despite the fact that you're miles away from me (and Hitler)? Or will the results now be guaranteed to be the same as your first observation? Has the probability waveform effectively been pre-collapsed before the experiment has even begun, turning a random event into a completely deterministic one?
Bonus question: What if you're actually out in space, exactly two light minutes away? You won't hear the results of my experiment until two minutes after I complete it (since it takes that long for my broadcast to reach you), so when you hear the results, you'll have to go back in time three and a half minutes in order to ensure you arrive before I actually begin. But then, if the results could be different this time, would that imply that information of your arrival in the past had reached me in only thirty seconds, despite you being two light minutes away?
My box is configured so that sixty seconds after I start the experiment, there will be a 50% chance that Hitler will be dead and a 50% chance that he'll be alive. At that point, I'm going to open the box and broadcast the results to the world.
...
...
...
He's alive! OK, that concludes that experiment.
But wait!, I imagine you saying. I wanted Hitler dead! Don't we all, my friend. Don't we all. But if you assume a model of time travel wherein the past can be changed, how about you just go back in time and so you can observe the experiment again?
So let's hypothetically say that you go back about a minute and a half, to before I've started the experiment. You don't interfere with the process in any direct way: you just quietly wait until it's completed and listen to the results again.
Will the experiment still be random, with a 50% chance of life or death? Does your mere presence in the past have the ability to alter the result, producing a potentially different answer this time around, despite the fact that you're miles away from me (and Hitler)? Or will the results now be guaranteed to be the same as your first observation? Has the probability waveform effectively been pre-collapsed before the experiment has even begun, turning a random event into a completely deterministic one?
Bonus question: What if you're actually out in space, exactly two light minutes away? You won't hear the results of my experiment until two minutes after I complete it (since it takes that long for my broadcast to reach you), so when you hear the results, you'll have to go back in time three and a half minutes in order to ensure you arrive before I actually begin. But then, if the results could be different this time, would that imply that information of your arrival in the past had reached me in only thirty seconds, despite you being two light minutes away?