3 ms·

This is actually incorrect. Pilot wave theory gives a deterministic explanation of phenomenon that the standard model treats as frequentist random events. Only

by kobeya 10y ago

This is actually incorrect. Pilot wave theory gives a deterministic explanation of phenomenon that the standard model treats as frequentist random events. Only if you assume no knowledge of the underlying hidden variables does the model reduce to the same as the frequentist standard model. This is exactly analogous to statistical mechanics (e.g. the ideal gas law) arising from classical Newtonian mechanics when you assume no specific knowledge about gas particles (individual positions or velocities), just aggregate mass and energy. An experiment that "proves" an extension of pilot wave theory might be, e.g., observing that decay rates of radioactive isotopes are affected by being in a quantum corral or something like that. An experiment that shows there is a more fundamental, and influenceable process underlying phenomenon that the standard interpretation treats as frequentist. For this reason it could be easily argued that theoretical physicists should be spending more time on pilot wave theory since it is this sort of theory from which future advances would come, vs the standard interpretation which puts up a semantic stop sign and says "proceed no further".

amluto 10y ago
> This is actually incorrect. Pilot wave theory gives a deterministic explanation of phenomenon that the standard model treats as frequentist random events. That's a remarkably incorrect statement. Let's see: * The "standard model" is a quantum field theory and has essentially nothing to do with this discussion. * Standard quantum mechanism says nothing about frequentist vs Bayesian interpretations of anything. * Standard quantum mechanism is a deterministic theory, too. Take a look at the Schrödinger equation: it's deterministic. Determinism is tricky. Suppose you have a particle doing its thing (going through double slits or whatever) and then you measure its position. In standard QM, you have two choices: you use the Born rule (or a fancier version -- I like the POVM formalism) and get a distribution of outcomes and a new post-measurement wavefunction, or you can include the measurement apparatus in your system and you end up with many worlds. In PWT, I think you have two similar choices. You can start with a distribution of your (hidden) initial condition and get an resulting distribution over the outcomes (which will be the same as the Born rule distribution if you calculate it correctly). If you do this, you still need to model decoherence, and I believe it looks at least as artificial as the standard QM case. Alternatively, you could presumably add degrees of freedom and more hidden variables to describe the measurement apparatus and you end up with an IMO extremely messy version of many worlds. <aside>My personal view is that standard QM with many worlds is the best theory we have. If you start with the many worlds hypothesis and assume that it's totally impractical to devise an experiment that actually detects coherence between large-scale states, you can ignore this coherence, average it out, recover the Born rule, etc, as an approximation. IOW I think that decoherence is a limitation of our ability to build crazy experiments rather than a fundamental property of the universe.</aside>

by kobeya 10y ago

amluto 10y ago

> This is actually incorrect. Pilot wave theory gives a deterministic explanation of phenomenon that the standard model treats as frequentist random events. That's a remarkably incorrect statement. Let's see: * The "standard model" is a quantum field theory and has essentially nothing to do with this discussion. * Standard quantum mechanism says nothing about frequentist vs Bayesian interpretations of anything. * Standard quantum mechanism is a deterministic theory, too. Take a look at the Schrödinger equation: it's deterministic. Determinism is tricky. Suppose you have a particle doing its thing (going through double slits or whatever) and then you measure its position. In standard QM, you have two choices: you use the Born rule (or a fancier version -- I like the POVM formalism) and get a distribution of outcomes and a new post-measurement wavefunction, or you can include the measurement apparatus in your system and you end up with many worlds. In PWT, I think you have two similar choices. You can start with a distribution of your (hidden) initial condition and get an resulting distribution over the outcomes (which will be the same as the Born rule distribution if you calculate it correctly). If you do this, you still need to model decoherence, and I believe it looks at least as artificial as the standard QM case. Alternatively, you could presumably add degrees of freedom and more hidden variables to describe the measurement apparatus and you end up with an IMO extremely messy version of many worlds. <aside>My personal view is that standard QM with many worlds is the best theory we have. If you start with the many worlds hypothesis and assume that it's totally impractical to devise an experiment that actually detects coherence between large-scale states, you can ignore this coherence, average it out, recover the Born rule, etc, as an approximation. IOW I think that decoherence is a limitation of our ability to build crazy experiments rather than a fundamental property of the universe.</aside>