Assuming collapse is quantitatively unlike assuming the existence of God. The collapse postulate is extremely unlikely a priori, in terms of the usual probabilities humans deal with, but the order of magnitude of unlikeliness is not even in the same order of magnitude as for an intelligent creator. Collapse is easy to describe (if you are careful and use thermodynamic degrees of freedom rather than consciousness to check when it will happen) and consistent with all observations. Of course MWI is way simpler, so I would bet on it against overwhelming odds; so would Scott Aaronson. That said, we still have huge unexplained things (the Born probabilities; the coexistence of GR and quantum) whose improbability a priori is very large compared to the gap between many worlds and collapse. So it is conceivable that we could discover a formulation with collapse which accounts for the Born probabilities and which is simpler than any formulation without collapse. I would bet anything against this, but it is certainly more likely than discovering that the world is flat or that God exists.
Human experience with quantum mechanics has been confined to experiments with extremely low entanglement. The question we are asking is: do the laws of physics we have observed so far continue to hold up in the high entanglement limit? This is very much analogous to asking: do the laws of physics we have observed so far continue to hold up in the high energy limit? Even if you expect the answer to be yes, there is broad consensus among physicists that investigating the question is worthwhile. I also suspect that if you knew more physics you would be less confident in your position.
(I recently took a class from Scott Aaronson which caused me to stop being confused about quantum mechanics very effectively and made me understand precisely how obvious MWI is. I think that his pedagogy is one of the most powerful forces in the world right now for getting students to understand quantum mechanics and see that MWI is obvious. I think you are basically manufacturing disagreement, which you can do successfully mostly because Scott Aaronson expresses himself less precisely than you do.)
Assuming collapse is quantitatively unlike assuming the existence of God.
Weird, because assuming the MWI is quantitatively like assuming atheism.
Your posts seems to suggest Scott Aaronson agrees with the MWI, but the blog post and Eliezer’s post seem to suggest he doesn’t—but your post also suggests Eliezer’s post is not evidence in this case. I am confused on the matter of what Scott actually believes.
MWI is qualitatively like atheism, but the weight of evidence on your aside (as compared to the next fundamentally different alternative) is so dissimilar that you should have qualitatively different beliefs about what updates may occur in the future. I guess there is also a qualitative difference; there are very many hypotheses as simple or simpler than theism which are equally “consistent” with observation, while there are relatively few as simple or simpler than collapse. Again, this is not to say that MWI isn’t overwhelmingly likely, but if we were to discover some evidence that its predictions stopped working in the high entanglement limit, some form of collapse is a reasonably likely departure from the simplest model. If we were to discover some evidence for say the intelligent design of life on Earth, a Christian god is still a spectacularly unlikely departure from the simplest model.
I don’t feel that the weight of the evidence is much different. For both atheism and MWI, I basically have Occam’s Razor plus some stretching on part of the opposing theories. “The way the world is underneath our observations” is really hard to get any hard evidence on.
Occam’s Razor doesn’t just tell you which option is more likely—it tells you how much more likely it is. The hypothesis that God exists is way, way more complicated than the alternative (I would guess billions or trillions of bits more complicated). The hypothesis that collapse occurs is more complicated than the alternative, but the difference isn’t in the same ballpark. Maybe you doubt that it is possible to concisely describe collapse, in which case we just have a factual disagreement.
Assuming collapse is quantitatively unlike assuming the existence of God. The collapse postulate is extremely unlikely a priori, in terms of the usual probabilities humans deal with, but the order of magnitude of unlikeliness is not even in the same order of magnitude as for an intelligent creator. Collapse is easy to describe (if you are careful and use thermodynamic degrees of freedom rather than consciousness to check when it will happen) and consistent with all observations. Of course MWI is way simpler, so I would bet on it against overwhelming odds; so would Scott Aaronson. That said, we still have huge unexplained things (the Born probabilities; the coexistence of GR and quantum) whose improbability a priori is very large compared to the gap between many worlds and collapse. So it is conceivable that we could discover a formulation with collapse which accounts for the Born probabilities and which is simpler than any formulation without collapse. I would bet anything against this, but it is certainly more likely than discovering that the world is flat or that God exists.
Human experience with quantum mechanics has been confined to experiments with extremely low entanglement. The question we are asking is: do the laws of physics we have observed so far continue to hold up in the high entanglement limit? This is very much analogous to asking: do the laws of physics we have observed so far continue to hold up in the high energy limit? Even if you expect the answer to be yes, there is broad consensus among physicists that investigating the question is worthwhile. I also suspect that if you knew more physics you would be less confident in your position.
(I recently took a class from Scott Aaronson which caused me to stop being confused about quantum mechanics very effectively and made me understand precisely how obvious MWI is. I think that his pedagogy is one of the most powerful forces in the world right now for getting students to understand quantum mechanics and see that MWI is obvious. I think you are basically manufacturing disagreement, which you can do successfully mostly because Scott Aaronson expresses himself less precisely than you do.)
Weird, because assuming the MWI is quantitatively like assuming atheism.
Your posts seems to suggest Scott Aaronson agrees with the MWI, but the blog post and Eliezer’s post seem to suggest he doesn’t—but your post also suggests Eliezer’s post is not evidence in this case. I am confused on the matter of what Scott actually believes.
MWI is qualitatively like atheism, but the weight of evidence on your aside (as compared to the next fundamentally different alternative) is so dissimilar that you should have qualitatively different beliefs about what updates may occur in the future. I guess there is also a qualitative difference; there are very many hypotheses as simple or simpler than theism which are equally “consistent” with observation, while there are relatively few as simple or simpler than collapse. Again, this is not to say that MWI isn’t overwhelmingly likely, but if we were to discover some evidence that its predictions stopped working in the high entanglement limit, some form of collapse is a reasonably likely departure from the simplest model. If we were to discover some evidence for say the intelligent design of life on Earth, a Christian god is still a spectacularly unlikely departure from the simplest model.
I don’t feel that the weight of the evidence is much different. For both atheism and MWI, I basically have Occam’s Razor plus some stretching on part of the opposing theories. “The way the world is underneath our observations” is really hard to get any hard evidence on.
Occam’s Razor doesn’t just tell you which option is more likely—it tells you how much more likely it is. The hypothesis that God exists is way, way more complicated than the alternative (I would guess billions or trillions of bits more complicated). The hypothesis that collapse occurs is more complicated than the alternative, but the difference isn’t in the same ballpark. Maybe you doubt that it is possible to concisely describe collapse, in which case we just have a factual disagreement.