3 ms·

So many-worlds says that the waveform doesn't collapse to a kronecker-delta, but to an average of multiple ones that 'don't interact'. Is this conjecture, do w

by rocqua 2y ago

So many-worlds says that the waveform doesn't collapse to a kronecker-delta, but to an average of multiple ones that 'don't interact'. Is this conjecture, do we see this in some specific toy systems, or do we know this is how all 'collapsed' waveforms behave? In other words, to what degree is the many-worlds interpretation just an interpretation of the waveform, or is it also a conjecture about how the waveform behaves? edit: it also sounds like the 'many worlds' are separate, but the 'real' waveform actually still has the information of all the worlds. That suggests the information in that waveform is huge. It also feels like the 'non-collapsed' things from the different worlds could still interact. How much of what I said here is wrong (I presume it is a lot)?

rcxdude 2y ago
It's a bit of both: it's an extrapolation from the results of smaller systems (backed up by experiments like the delayed-choice quantum eraser, which creates an 'observer' which is in a superposition with respect to the experimenter-observer), but it's also reasonably well backed up by theoretical results that large, hot systems will tend to decohere or separate (the result in this paper is interesting because it shows that for this kind of state there's a hard shutoff where quantum effects completely disappear, instead of things just getting less and less 'blurry') And yes, in many ways the conjectured real waveform is quite a beast. Though you could also argue that the information content is actually very low, in that it is fully specified by the boundary condition of the big bang (well, at least any quantum-like correct theory of everything, remember gravity isn't explained by QM). Thus it's any given universe that actually represents a lot of information because it's the result of many different 'decisions' as that initial state starts to become very complex as it evolves (in a similar way to which pi can be simply defined but contains infinite digits, and if it's 'normal', you can find any string of digits in there, but it'll almost certainly take more digits to specify the position of the sequence in pi than are in the sequence). There's often similar objections on the basis of energy/mass which can be addressed in a similar way. And, as far as I understand it the interactions between 'worlds' can only really happen for things that are 'close by', in terms of how much the states have separated (that is, in order to observe quantum effects you need to have isolated the system well enough it hasn't already entangled with you. Once it has you can't undo that). I think one general challenge with conceptualizing it is that in practice the correlations in the waveform are extremely messy: the whole universe is not entangled all at once. The 'splits' suggested in the interpretation are basically creating a potentially infinite amount of universes every time any entanglement occurs, which makes it really hard to reason about (indeed, there are philosophical issues that arise if you try to imagine what you should expect to see as a subjective observer in such a universe)

by rocqua 2y ago

rcxdude 2y ago

It's a bit of both: it's an extrapolation from the results of smaller systems (backed up by experiments like the delayed-choice quantum eraser, which creates an 'observer' which is in a superposition with respect to the experimenter-observer), but it's also reasonably well backed up by theoretical results that large, hot systems will tend to decohere or separate (the result in this paper is interesting because it shows that for this kind of state there's a hard shutoff where quantum effects completely disappear, instead of things just getting less and less 'blurry') And yes, in many ways the conjectured real waveform is quite a beast. Though you could also argue that the information content is actually very low, in that it is fully specified by the boundary condition of the big bang (well, at least any quantum-like correct theory of everything, remember gravity isn't explained by QM). Thus it's any given universe that actually represents a lot of information because it's the result of many different 'decisions' as that initial state starts to become very complex as it evolves (in a similar way to which pi can be simply defined but contains infinite digits, and if it's 'normal', you can find any string of digits in there, but it'll almost certainly take more digits to specify the position of the sequence in pi than are in the sequence). There's often similar objections on the basis of energy/mass which can be addressed in a similar way. And, as far as I understand it the interactions between 'worlds' can only really happen for things that are 'close by', in terms of how much the states have separated (that is, in order to observe quantum effects you need to have isolated the system well enough it hasn't already entangled with you. Once it has you can't undo that). I think one general challenge with conceptualizing it is that in practice the correlations in the waveform are extremely messy: the whole universe is not entangled all at once. The 'splits' suggested in the interpretation are basically creating a potentially infinite amount of universes every time any entanglement occurs, which makes it really hard to reason about (indeed, there are philosophical issues that arise if you try to imagine what you should expect to see as a subjective observer in such a universe)