You bounce a photon off a half-silvered mirror and don’t look at the results: no universe split.
You bounce a photon off a half-silvered mirror and look at the results: Bam! Split universe.
MWI:
You bounce a photon off a half-silvered mirror and don’t look at the results. Since the physical state of your brain is not causally dependent on the destination of the photon, you don’t branch into two mwenglers in any noticeable way.
You bounce a photon off a half-silvered mirror and look at the results. Since you’ve made the state of your brain causally dependent on an event with quantum randomness, you branch into two mwenglers which are different on a macroscopic level. Two persons, which happen to share a causal history up to looking at the experimental outcome.
… You bounce a photon off a half-silvered mirror and look at the results: Bam! Split universe.
Copenhagen Interpretation never splits universes. Instead, you have a wave function collapse in the one and only universe.
You bounce a photon off a half-silvered mirror and don’t look at the results. Since the physical state of your brain is not causally dependent on the destination of the photon, you don’t branch into two mwenglers in any noticeable way.
In MWI, you NEVER branch in to two anythings in a “noticeable” way. All the myriads of branches have no interactions, there is nothing noticeable about any of the other branches from within the branch we are in. If there is something noticeable about other branches, then an experiment could be defined to check the hypothesis of branching, and we would start to gather evidence for or against branching. Until such time as an hypothesis is created and tested and shows evidence for branches, MWI is an interpretation, and not a theory.
So why does it even matter? I am thinking it through and realizing that an interpretation is in some way a pre-theory. As we sit with the idea of MWI, maybe one of us develops hypotheses about experiments whic might show evidence for the other branches, or not. Without the interpretation of MWI, that hypothetical progress might never be available.
They do interact. This is how quantum physics was discovered.
The problem is that the magnitude of interaction is getting very small very quickly, so after a few microseconds it becomes technically impossible to measure. This is what allows people to say: “Yeah, for a few microseconds there is something mathematically equivalent to branches, but then it disappears completely” and you can’t experimentally prove them wrong.
One side believes that the interaction is getting smaller, but it never reaches exactly zero. Other side believes that the interaction is getting smaller, and then in some unspecified moment all branches except one disappear. Experimental data say that the interaction is getting smaller until it becomes too small to see… and then, well, it is too small to see what happens. So essentially both sides disagree about who has the burden of proof; about the exact meaning of “fewest assumptions” in Occam’s razor. One side says that “the extra branches disappearing” is the extra assumption. Other side says that “the extra branches not disappearing, even when their interaction becomes too small to measure” is the extra assumption.
More precisely, the magnitude of interaction depends on how much the particles in the branches are different. Therefore the branches we have measurable interaction with are those almost the same as our branch. The interaction is largest when both branches are exactly alike except for one particle. This is the famous double-slit experiment—two branches of the universe with the particle going through different slits interact with each other. The branches are there. The question is not whether multiple branches exist, but whether they disappear later when their interaction becomes very small.
maybe one of us develops hypotheses about experiments which might show evidence for the other branches, or not.
How do you prove experimentally that the other branches do not disappear, especially if your opponents refuse to specify when they should disappear. If you make an experiment that proves that “after N seconds, the branches still exist”, your opponents can say: “Yeah, but maybe after N+1 seconds they disappear.” Repeat for any value of N.
Copenhagen:
You bounce a photon off a half-silvered mirror and don’t look at the results: no universe split.
You bounce a photon off a half-silvered mirror and look at the results: Bam! Split universe.
MWI:
You bounce a photon off a half-silvered mirror and don’t look at the results. Since the physical state of your brain is not causally dependent on the destination of the photon, you don’t branch into two mwenglers in any noticeable way.
You bounce a photon off a half-silvered mirror and look at the results. Since you’ve made the state of your brain causally dependent on an event with quantum randomness, you branch into two mwenglers which are different on a macroscopic level. Two persons, which happen to share a causal history up to looking at the experimental outcome.
Copenhagen Interpretation never splits universes. Instead, you have a wave function collapse in the one and only universe.
In MWI, you NEVER branch in to two anythings in a “noticeable” way. All the myriads of branches have no interactions, there is nothing noticeable about any of the other branches from within the branch we are in. If there is something noticeable about other branches, then an experiment could be defined to check the hypothesis of branching, and we would start to gather evidence for or against branching. Until such time as an hypothesis is created and tested and shows evidence for branches, MWI is an interpretation, and not a theory.
So why does it even matter? I am thinking it through and realizing that an interpretation is in some way a pre-theory. As we sit with the idea of MWI, maybe one of us develops hypotheses about experiments whic might show evidence for the other branches, or not. Without the interpretation of MWI, that hypothetical progress might never be available.
They do interact. This is how quantum physics was discovered.
The problem is that the magnitude of interaction is getting very small very quickly, so after a few microseconds it becomes technically impossible to measure. This is what allows people to say: “Yeah, for a few microseconds there is something mathematically equivalent to branches, but then it disappears completely” and you can’t experimentally prove them wrong.
One side believes that the interaction is getting smaller, but it never reaches exactly zero. Other side believes that the interaction is getting smaller, and then in some unspecified moment all branches except one disappear. Experimental data say that the interaction is getting smaller until it becomes too small to see… and then, well, it is too small to see what happens. So essentially both sides disagree about who has the burden of proof; about the exact meaning of “fewest assumptions” in Occam’s razor. One side says that “the extra branches disappearing” is the extra assumption. Other side says that “the extra branches not disappearing, even when their interaction becomes too small to measure” is the extra assumption.
More precisely, the magnitude of interaction depends on how much the particles in the branches are different. Therefore the branches we have measurable interaction with are those almost the same as our branch. The interaction is largest when both branches are exactly alike except for one particle. This is the famous double-slit experiment—two branches of the universe with the particle going through different slits interact with each other. The branches are there. The question is not whether multiple branches exist, but whether they disappear later when their interaction becomes very small.
How do you prove experimentally that the other branches do not disappear, especially if your opponents refuse to specify when they should disappear. If you make an experiment that proves that “after N seconds, the branches still exist”, your opponents can say: “Yeah, but maybe after N+1 seconds they disappear.” Repeat for any value of N.