which should win?

[MorpheusSandman]

Looking past the fact that Eliezer agrees with you, can you recommend his article(s) just on the basis of content minus the sales job, or (preferably) do you have a more neutral source of information?

That Yudkowsky article is the final one in an extremely long sequence that begins here. When I first started reading about QM many years ago, I knew a QM mechanics professor online who actually recommended Yudkowsky as a good, free, accurate, online source. While this professor disagreed somewhat with Yudkowsky's confidence in the MW conclusion, he did claim the actual information was accurate. In my subsequent years of reading about QM and talking about it with others even more knowledgeable than myself, I've yet to see anyone note any error in that sequence, even if some do disagree with the "rhetoric." That "conclusion" article I linked to is mostly rhetoric because Yudkowsky is trying to condense all of the information from the sequence and show how it argues convincingly for MW. Most any conclusion from any non-fiction book contains some kind of rhetorical point for which readers are supposed to have the background information in their mind. I just quoted that part because I think Yudkowsky gives a convincing argument as to why the case is so bad for single world interpretations; but, yeah, you kinda have to understand something about QM to even know what he's referring to.

It also helps to have some background in a few other things to get where Yudkowsky is coming from, especially his version of epistemic rationality informed by Bayes' Theorem. The thing is, there are currently no ways to "test" between CI and MW, that's why they remain "interpretations" and not "theories." However, for Yudkowsky the preference for CI amongst most physicists (though this preference depends on who, precisely, you ask; it seems that MW is just as, if not more, preferred amongst the top-level physicists) seems to highlight for Yudkowsky several fundamental flaws in human rationality, even amongst scientists. For one it seems that even scientists are slow to let go of certain intuitions like there being a "single world," preferring "magical" explanations like the various collapse interpretations (CI is just the most famous) that unnecessarily complicates QM and creates all of these seemingly irresolvable problems. Why do they do this? Probably because one world seems "less weird," and because the collapse interpretations came first.

However, remove our (infamously faulty) intuitions, consider the current evidence on its face, and then apply mathematical logic, there's no question that MW is the preferable interpretation. If you'd like, I could try to explain it to you myself. I usually get compliments on my abilities to relate complicated, esoteric subjects to laymen, and while I would not claim to be an expert, I do think I've read more than enough on this subject to lay out the basic issues and the conflict behind CI and MW. I can do this without making a "rhetorical" case for MW. I will simply explain the fundamental issue, what CI says about it, what MW says about it, and then lay out the problems and evidence with/for both.

Yes, it's pure rhetoric. And he calls himself "The Physicist"! Read *actual* books by top physicists, with *actual* names, working at top universities, like Stephen Hawking and Brian Greene. Otherwise how do you know you aren't reading stuff made up by a janitor?

Balderdash.

1. "Actual books" by "top physicists" are not usually free of rhetoric, because even scientists have their preferred theories/interpretations for which there is no current proof.

2. "Actual books" by "top physicists" are not necessarily the best starting points for laymen, because "top physicists" get to be "top physicists" by impressing other "top physicists" and not necessarily enlightening laymen as to what the hell all these "top physicists" are talking about. Popularizing science is a completely different skill that most "top scientists" don't possess.

3. Based on all my reading of "top physicists," the actual information (ignoring the rhetorical arguments) in both articles I linked to are accurate. Are there arguments to be made against the rhetoric? Yes. In fact, there is one argument on the very site I culled that long quote from here. It's also notable that there are lots of debates going on in the TALK sections of these articles. So, lest anyone feel the rhetoric is too one-side, it's really not.

4. I assume your "janitor" question was a bit of rhetoric itself, though it's easily dismissed the moment one realizes that almost no janitors would even know the terminology being used in these articles and, what's more, one can relatively easy check the information given amongst any number of sources both online and off. It would be one thing for you to claim some kind of authority and then quibble with the actual information or arguments, but you're just committing a blatant reverse Argument from Authority fallacy and have not given one reason to discredit either article or author I linked to.

Those many worlds are just as unfalsifiable as any deity.

You've said this many times, I've countered many times, you've ignored many times. The "many worlds" themselves are a consequence of assuming QM works all the way down. The assumption IS falsifiable, and if the assumption is falsifiable, then so are the many worlds. The problem is not that MW is unfalsifiable (it is), the problem is that it hasn't been falsified. A deity is not a logical consequence of any other assumption that is likewise falsifiable.

You mentioned in that post that "CI treats the wavefunction as non-real". That is to its credit.

The problem is that if wavefunctions were non-real then we shouldn't be able to use them to make predictions. Why? Because if they were just possibilities before observed then they should not translate into predictions prior to observance, yet they do. Cioran tried to explain this in our first thread on the matter. What's more, why in the world is it preferable to assume something for which we see direct evidence for (how particles behave in a double-slit experiment) as NON-REAL? How can non-real entities provide such evidence for themselves?

I don't see how that could be evidence for Many Worlds.

Because there's no reason whatsoever to assume the wavefunction is non-real given that we can see the direct effects of it and make predictions based on it. How does a non-real entity do that?

I don't think the only indeterminism that exists is at the quantum level.

All of physics and even your chosen QM interpretation disagrees with you.

1) Deterministic and Local: This position has been falsified by science (Bell's theorem and the experiments confirming it). It seems as if Many Worlds proponents think they can recover it. This makes Many Worlds unscientific. Failure to derive the Born probabilities confirms that MW is not even an interpretation for QM.

This position was falsified by assuming that human observers are outside and separate from the very processes they're observing and, what's more, it assumes the existence of the collapse whose appearance is explained by MW. So, that position was only falsified if we assume the very things that MW questions. Try again.

2) Indeterministic and Local: This is where the Copenhagen interpretation as well as the various alternate, consistent or decoherent histories interpretations fit. This is the only position for which there is evidence for matter-energy entities.

Proof you don't even understand your own chosen interpretation. CI is non-local, and can only get around this non-locality by assuming the wavefunction is nonreal; but assuming the wavefunction is nonreal doesn't suddenly make CI local, because you still have the problem of simultaneous measurements affecting others at great distances at millions of times faster than the speed of light. Yudkowsky gives one example in the part of the article I quoted.

What's more, there is no "evidence" for Copenhagen. There is an observation (probabilistic wavefunctions turn into particles when we observe them) and an assumption made on that observation (observation causes collapse) for which there is no evidence either in the testing or the math, and, what's more, it's an interpretation that conflicts with everything else we know about physics and for which MW actually explains while being reconcilable with everything else we know about physics. Simply put: there is NO EVIDENCE WHATSOEVER FOR A COLLAPSE IN THE MATH OR EXPERIMENTS.

[mal4mac]

Balderdash.

1. "Actual books" by "top physicists" are not usually free of rhetoric, because even scientists have their preferred theories/interpretations for which there is no current proof.

You have a point, but they usually are less rhetorical, and present the other side. If they do not do this they would get into a great deal of trouble from their colleagues, and if they want to keep their reputation as physicists they have to present a balanced overview. Here's an example from a top physicist, Sean Carroll, who actually prefers the many worlds interpretation, but presents a balanced perspective, even highlighting that more physicists prefer the Copenhagen interpretation:

http://www.youtube.com/watch?v=ZacggH9wB7Y

He slips into rhetoric, for a moment, when he suggests that his fellow physicists might not have thought through the issues as completely as he has. (His Copenhagen pals would just smile at that, and accuse him of being a "bit naughty", knowing fine well that this is rhetoric!) But note, he's careful not to say his fellow physicists are "just wrong", because they aren't, and he doesn't want to be thought of as a crank.

2. "Actual books" by "top physicists" are not necessarily the best starting points for laymen, because "top physicists" get to be "top physicists" by impressing other "top physicists" and not necessarily enlightening laymen as to what the hell all these "top physicists" are talking about. Popularizing science is a completely different skill that most "top scientists" don't possess.

That has not been my impression; Albert Einstein, Stephen Hawking, Steven Weinberg, Richard Feynman and Brian Greene are probably the best popularisers I've read, and they are certainly top physicists! The average science journalist just doesn't know enough physics, and are always getting things wrong, so you can never know if you are getting the correct overview unless it comes straight from the top guys.

It would be one thing for you to claim some kind of authority and then quibble with the actual information or arguments, but you're just committing a blatant reverse Argument from Authority fallacy and have not given one reason to discredit either article or author I linked to.

So you're happy reading any old web page rather than Richard Feynman? Next time you are ill will you consult an authorised doctor, or use Google and take the advice of "Mr Doctor" on the first page you see? From what you have said previously, you aren't a physicist, so your link is a case of the blind leading the blind up a blind alley.

[MorpheusSandman]

You have a point, but they usually are less rhetorical, and present the other side.

Both of those articles present the other side, they just present it in an unflattering (but not inaccurate) light. It is a bit ridiculous that you have these two "competing" interpretations, one which has to make all kinds of unnecessary, complicating assumptions only to reach a point where the answer inexplicably conflicts with everything we know about physics (Copenhagen); and another for which we only have to take what we see at face value and work out the issue from there, and from that we reach a point of many worlds. To want both sides presented as being "equally" likely is like the "rhetoric" I see from Creationists/IDers who want their theories presented as being "equal" with evolution even when there is no equality in the evidence. It's true that the CI/MW issue isn't THAT severe, but the more one looks at it, it's not dissimilar. I think both of the articles I linked to accurately reflect WHY these two interpretations don't deserve equal consideration.

That has not been my impression; Albert Einstein, Stephen Hawking, Steven Weinberg, Richard Feynman and Brian Greene are probably the best popularisers I've read, and they are certainly top physicists!

For every top scientist who's also a great popularizer you could name there are probably 50-100 top scientists who could no more popularize than I could eat a crocodile.

So you're happy reading any old web page rather than Richard Feynman?

What makes you think I haven't read both? I started with Yudkowsky; I've moved on; I still think Yudkowsky is as good an intro as there is.

However, you shouldn't mistake professionalism with actual knowledge. My profession is online poker; there are lots of non-poker players that know far more about the game than I do (they teach/write books about the game). Similarly, I am not a film professor, yet I know far more about film than a great many films professors because I've spent more time studying film than most film professors do. One of the perks of playing professional poker is I have lots of free time, because I don't use my profession to live a lavish lifestyle, but to allow me to indulge in my other non-expensive passions like reading, writing, learning, etc.

Yudkowsky is, by profession, an AI Researcher. His profession has lead him into a great many related but tangential areas including QM. The fact that no actual physicist I've talked to has noted a mistake in his series, along with the fact I haven't read anything from any actual physicist, like Feynman, that contradicts anything in his series, is good enough for me that it's accurate.

It doesn't hurt that I also subscribe to Yudkowsky's philosophy of epistemic rationality based on Bayes' Theorem. Most don't even know what those things are so many of his posts come off as being one-sided rhetoric. If people were perfectly rational there would be no need for rhetoric. The rhetoric exists because people aren't rational. In fact, LessWrong is primarily a blog about refining the art of rationality, which basically requires unlearning almost every instinct/intuition you have about life/reality and becoming aware of the manifold cognitive biases that afflict human brains. If you did that, and read through the QM sequence, you'd come to the conclusion that MW was the best interpretation without the need for any rhetoric.

Here's an example from a top physicist, Sean Carroll, who actually prefers the many worlds interpretation, but presents a balanced perspective, even highlighting that more physicists prefer the Copenhagen interpretation:

http://www.youtube.com/watch?v=ZacggH9wB7Y

I watched this video. Besides Carroll's affability, which may give the impression of him presenting a "balanced perspective," I don't really see anything in that video that's balanced. He gives the poll data, he goes on to explain MW, to highlights some reasons why he prefers it (it's simpler, resolves itself with classic physics, removes the unnecessary assumptions of CI); then he talks about the poll data and makes the basic claim both Yudkowsky and the "Ask a Physicist" physicist make: that CI makes no sense and is only accepted by people who don't think about it. So, to repeat, I'm not sure what you see as "balanced" about the video...

PS, I liked Carroll's analogy about KNOWING VS DOING (race car drivers VS mechanics) and how that helps explain the poll data. I was making essentially the same point above regarding professionalism VS knowledge. Lots of professionals know less about their profession than those that simply write about that profession. Being a physicist and knowing physics are not necessarily the same thing.

[mal4mac]

... CI makes no sense and is only accepted by people who don't think about it.

He didn't say that!

[YesNo]

The problem is that if wavefunctions were non-real then we shouldn't be able to use them to make predictions. Why? Because if they were just possibilities before observed then they should not translate into predictions prior to observance, yet they do. Cioran tried to explain this in our first thread on the matter. What's more, why in the world is it preferable to assume something for which we see direct evidence for (how particles behave in a double-slit experiment) as NON-REAL? How can non-real entities provide such evidence for themselves?

If you construct a Bayes formula to predict the probabilities, is that formula a real entity or field? Or is it just a formula? I see the wave function as a formula to compute probabilities and it predicts the way a group of these particles will behave. It does not predict the way any individual one of them will behave.

I can see having a field of forces such as the electromagnetic field, but I find it hard seeing a field of probabilities.

This position was falsified by assuming that human observers are outside and separate from the very processes they're observing and, what's more, it assumes the existence of the collapse whose appearance is explained by MW. So, that position was only falsified if we assume the very things that MW questions. Try again.

Are you claiming that MW will bring back both determinism and localism?

Proof you don't even understand your own chosen interpretation. CI is non-local, and can only get around this non-locality by assuming the wavefunction is nonreal; but assuming the wavefunction is nonreal doesn't suddenly make CI local, because you still have the problem of simultaneous measurements affecting others at great distances at millions of times faster than the speed of light. Yudkowsky gives one example in the part of the article I quoted.

What's more, there is no "evidence" for Copenhagen. There is an observation (probabilistic wavefunctions turn into particles when we observe them) and an assumption made on that observation (observation causes collapse) for which there is no evidence either in the testing or the math, and, what's more, it's an interpretation that conflicts with everything else we know about physics and for which MW actually explains while being reconcilable with everything else we know about physics. Simply put: there is NO EVIDENCE WHATSOEVER FOR A COLLAPSE IN THE MATH OR EXPERIMENTS.

At one point I thought the experiments verifying that Bell's inequality failed implied that quantum reality was nonlocal, but I don't think that is the case anymore. However, that would depend on there being evidence that matter-energy entities can interact non-locally. Do you have such evidence? If so, I would have to move CI to the Indeterminism and NonLocal category.

I agree with mal4mac's suspicions about Yudkowsky. If you would like an atheistic physicist to provide an alternate description of QM and its interpretations, I suggest Victor Stenger, "The Unconscious Quantum", 1995. When he is discussing physics, I find him interesting. When he goes into his anti-religious ranting, I ignore him.

[MorpheusSandman]

He didn't say that!

He says, paraphrased, that he thinks the majority accept CI because they haven't thought about it very hard. Of course, he lightens and qualifies this by noting it's his personal opinion (that he believes is true), but that's essentially what he says. In fact, he makes it even clearer when he says, again paraphrased: "that's not to say if you think about it very hard you are lead to MW, but you are lead away from CI. The more people worry about how QM works, the less likely they are to accept CI."

What Carroll is really doing is making the distinction between applied physicists and theoretical physicists. Applied physicists don't really care about which interpretation is right, so they don't think about it and merely accept CI because it's, as Carroll said, the one that's most widely taught in the textbooks. Most applied physicists say "shut up and calculate" because what interpretation is right has zero impact on their ability to manipulate QM to practical uses (like, say, in GPS). I suspect that if you polled theoretical physicists only the poll results would be very different, and that's what Carroll means by people who really think about this stuff. Most physicists don't because that's not what they're paid to do.

If you construct a Bayes formula to predict the probabilities, is that formula a real entity or field? Or is it just a formula? I see the wave function as a formula to compute probabilities and it predicts the way a group of these particles will behave. It does not predict the way any individual one of them will behave.

The numbers we apply to Bayes' are obtained by our observations of real objects/events, just like the formulas we've obtained from QM we obtained by observing how wavefunctions behave. I can't think of any formulas in any science that refer to nonreal things. You say that you "see" the wavefunction as a formula, yet how could we have obtained this formula without the wavefunction being real to begin with? How can you see the effects of something nonreal?

Are you claiming that MW will bring back both determinism and localism?

This is me facepalming... MW already has "brought back" (LOL, where did they go?) both determinism and localism. By assuming QM works all the way down (or up), QM becomes local and determinate.

However, that would depend on there being evidence that matter-energy entities can interact non-locally. Do you have such evidence? If so, I would have to move CI to the Indeterminism and NonLocal category.

The EPR Paradox proved non-locality under CI, but assumed locality under the assumption of hidden variables; Bell disproved the possibility of any variables that could account for the paradox. So, between EPR and Bell you have the proof of nonlocality under CI. Besides the disproved hidden variables, there's no way to get around non-locality in CI. There have been many Bell-like experiments done: http://en.wikipedia.org/wiki/Bell_te...le_experiments

However, all of those experiments only prove non-locality under the assumption of collapse. That's the key word/term: ASSUMPTION of collapse. All of those experiments place the observer (us) outside the observed. They in no way conflict with MW. In fact, only MW can explain why we see what we do.

I agree with mal4mac's suspicions about Yudkowsky.

Yet can not point to anything in any of his articles that's actually wrong.

[Calidore]

That Yudkowsky article is the final one in an extremely long sequence that begins here.

Ah. Yes, I would rather start at the beginning than the end, thanks.

However, remove our (infamously faulty) intuitions, consider the current evidence on its face, and then apply mathematical logic, there's no question that MW is the preferable interpretation. If you'd like, I could try to explain it to you myself. I usually get compliments on my abilities to relate complicated, esoteric subjects to laymen, and while I would not claim to be an expert, I do think I've read more than enough on this subject to lay out the basic issues and the conflict behind CI and MW. I can do this without making a "rhetorical" case for MW. I will simply explain the fundamental issue, what CI says about it, what MW says about it, and then lay out the problems and evidence with/for both.[/QUOTE]

I'd rather start at the very bottom, as I have zero interest in arguing one set of complete intangibles over the other. I've asked a couple of times (not of you, specifically) in other threads on this subject if anyone could lay out what's actually known, through verifiable and repeatable experimentation, what we have good reason to believe is true, and what's simply speculation. Or, as I also put it, where the tree trunk of knowledge ends and the branches of speculation begin. I'd also be interested in knowing what we'd need to convert the speculation into fact.

[YesNo]

The numbers we apply to Bayes' are obtained by our observations of real objects/events, just like the formulas we've obtained from QM we obtained by observing how wavefunctions behave. I can't think of any formulas in any science that refer to nonreal things. You say that you "see" the wavefunction as a formula, yet how could we have obtained this formula without the wavefunction being real to begin with? How can you see the effects of something nonreal?

The mathematics is just a tool to describe what is real. It is a map of the territory, not the territory. (I'm trying to quote something I think Yudkowsky wrote.) It's the menu, not the meal.

This is me facepalming... MW already has "brought back" (LOL, where did they go?) both determinism and localism. By assuming QM works all the way down (or up), QM becomes local and determinate.

So, overall the many worlds there is determinism and localism, but within each of the those worlds QM must work as it does in our world. Is that what MW claims? Since we can't ever get to a perspective where we can see this determinism and local behavior, what good is it?

The EPR Paradox proved non-locality under CI, but assumed locality under the assumption of hidden variables; Bell disproved the possibility of any variables that could account for the paradox. So, between EPR and Bell you have the proof of nonlocality under CI. Besides the disproved hidden variables, there's no way to get around non-locality in CI. There have been many Bell-like experiments done: http://en.wikipedia.org/wiki/Bell_te...le_experiments

However, all of those experiments only prove non-locality under the assumption of collapse. That's the key word/term: ASSUMPTION of collapse. All of those experiments place the observer (us) outside the observed. They in no way conflict with MW. In fact, only MW can explain why we see what we do.

Yes, I guess a sort of non-locality is present in those experiments, but I was thinking of a non-locality in which one can communicate non-locally, not just influence the results of measurements on the other side of the universe. But I guess that is non-local enough. I think it is sometimes referred to as non-separability. OK, I'll put CI in the Indeterminism and NonLocal category.

Yet can not point to anything in any of his articles that's actually wrong.

Since I don't think MW is true, he would be wrong about that. However, most people aren't wrong all the time.

[MorpheusSandman]

Ah. Yes, I would rather start at the beginning than the end, thanks.

No problem. I only quoted the end because it was directly pertinent to what I was talking about. Be forewarned, though, that the entire sequence does contain some maths that even I could barely follow.

I've asked a couple of times (not of you, specifically) in other threads on this subject if anyone could lay out what's actually known, through verifiable and repeatable experimentation, what we have good reason to believe is true, and what's simply speculation. Or, as I also put it, where the tree trunk of knowledge ends and the branches of speculation begin.

Oh, well, that's rather easy.

Facts: Particles appear to be in multiple positions at once before we measure them. We call this "superposition" or "the wavefunction." From this state, we can not measure precisely the position or momentum of a particle. We can measure one to a degree of accuracy, but the more accurate we get on position or momentum, the less accurate we get on the other. The mathematics that models the limits of our accuracy is known as The Uncertainty Principle.

What's more, when we make the measurement or observation this "superposition" or "wavefunction" collapses to a particle that then behaves deterministically according to the laws of General Relativity. The simplest and most common experiment done to prove this "superpositioned" state is the double-slit experiment. If you don't want to read the dry Wikipedia page, here's a short video for a Children's Education film that just lays out the experimental results.

Basically, all this boils down to this: When we aren't looking/measuring/observing, particles display characteristics of both particles and waves, at times going through both slits, neither slit, or just one or the other at apparent random. The equation that models this is the Schrodinger Equation.

The above are the basic, essential "facts" of what's known. There are certainly many more facts, but the above is essential first step. Most everything else proceeds from how we look at the above. One can certainly draw most of the fundamental questions from what's NOT known about the above:

1. How can particles before observation display characteristics of both discrete matter and waves?
2. What role does our observation/measurement play in making the wavefunction "collapse" into just a particle that behaves according to General Relativity?
3. What role does General Relativity, our most complete picture of physics on the macro scale, play in quantum mechanics, and why do GR and QM seem to contradict each other?
4. Given the both/neither/one-or-other nature of the wavefunction, is quantum mechanics local or non-local, deterministic or non-deterministic, real or non-real?

To explain the terms in 4., locality in physics is a basic principle that matter can only affect other matter locally and not at great distances. This is well-tested within General relativity, but under the classic view of QM this principle of locality seems to be violated. It gave rise to Einsten's phrase "spooky action at a distance." Determinism and indeterminism are more self-explanatory. Under General Relativity, physics is determinate, meaning that every movement of every physical object can be measured and predicted. QM, because of the unpredictability, seems to violate this, and be indeterministic. One way many interpretations get around this is to assume that the wavefunction is "non-real," meaning that it's only a formula of potential, and doesn't become "real" until observance.

Now for the two major "interpretations" or "speculation" as you would call them.

Copenhagen or CI for short: CI believes in the "split" between the quantum and macro world, that the quantum world behaves via one set of physical laws modeled mostly by Shrodinger and Heisenberg, and the macro world behaves by another, mostly modeled by General Relativity. Observation causes the "collapse" of the wavefunction to a particle. CI is non-local, meaning that particles can affect each other thousands of miles away at millions of times faster than the speed of light; it's non-real, meaning it treats the wavefunction as merely an abstract formula; it's indeterministic, meaning that we cannot predict the position/momentum of particles when they "collapse" to the single world.

Problems with CI: The biggest problem with CI is that it contradicts everything else we know about physics (especially via General Relativity). We also have no apparent reason for why observation/measurement should cause a particle collapse. So CI presents a view that there are these two worlds, the quantum and the macro, that refuse to meet in the middle. Indeed, we don't even know where that middle is; where, exactly, the quantum becomes the macro.

The Infamous Cat: Shrodinger's Cat is, basically, the "macro" illustration of CI. Before we open the box, the cat is both alive and dead (like a particle is both a particle and a wave); once we observe the cat it suddenly becomes alive OR dead, like how once we observe a wavefunction it becomes a particle and goes through only one slit.

Back to some facts: Disturbed by these issues, and before quantum non-locality was known for certain, Einstein, along with Podolsky and Rosen, proposed a thought experiment called the EPR Paradox that proved the non-locality under the collapse interpretation. However, assuming locality, Einstein used this paper to argue that there must be hidden, unknown variables that made QM seem non-local. Later, John Bell published what's known as Bell's Theorem which proved that no hidden variables could account for all of the predictions of QM. So this put an end to the possibility that something we didn't know could resolve all of these issues. QM is definitely non-local under the collapse interpretation.

A Preliminary with Many Worlds, or; A Rose By Any Other Name: I just wanted to note that it probably would've been better if MW had been named after its founder Hugh Everett, like Copenhagen was named after its founder. The term "MW" refers to a consequence of the interpretation, not the assumptions being made. Keep that in mind.

Now, back to the speculation itself:

Many Worlds: MW eliminates the collapse postulate and assumes that QM work all the way down (or up, depending on where you start). This means that the "macro" world we see is just the aggregate of a huge number of particles in certain configurations. MW is real, local, and deterministic. It's real because it assumes the equations that govern QM are themselves encoded into the wavefunction; it's local because it explains how measurements are actually two quantum systems (observer and observed) becoming entangled; it's deterministic because if the wavefunction formulas are real, then the wavefunction is just evolving linearly according to those equations.

With these assumptions, MW explains what is going on in the double-slit experiment like this: put in the language of Shrodinger's Cat, there is a world where the cat is dead and one where it's alive. These two worlds are "interfering" with each other. Once you open the door to look, you too "split" and see an "alive" cat in one world and a "dead" cat in another. However, these worlds are, essentially, already present. The wavefunction is evidence of them "interfering" with each other, and observance is what causes "decoherence" between worlds (rather than collapse). A better way to think of it is like this: you have two quantum systems in superposition (in Shrodinger's Cat those systems are you and the cat). Observaton causes these two systems to "entangle" with each other, and the superpositions "pair off" and "part ways" from their superpositioned state.

Problems with Many Worlds: There is only one problem with MW and it's this: MW can not derive the Born rule. The Born rule is the means by which we're able to make predictions using QM as it gives us the probabilities that any measurement will end up one way instead of another. Under MW, there's no apparent reason why one world would/should be more likely than another. Thus far, there's no explanation that resolves this (that I'm aware of, at least).

I should stress that the many worlds themselves ARE NOT A PROBLEM WITH MW. The correct way to assess the various QM interpretations is too look at what they assume and then look at what problems those assumptions create. Copenhagen assumes WF collapse into a single world; this creates the problem of it being irreconcilable with classic physics without disproving classic physics. MW assumes that everything is REALLY QM; this creates the problem of not being able to derive the Born rule. The actual "many worlds" are just a consequence of MW's assumptions being true. You could say it takes the facts of QM at face value, that particles are really going through both slits/neither slit/one-or-the-other slit because until we, as quantum systems ourselves, become entangled with the particles we can glimpse every possible "world" via the experiment's results.

Conclusions + Some Thoughts

So, one basic question you might ask yourself is this: Is it more weird that we have two contradictory, irreconcilable models for two different "levels" of reality without even knowing at what point one world "segues" into the other, or why our observation causes this "collapse" to begin with (Copenhagen); or is it more weird that all of reality, including ourselves, behaves according to the laws of QM, but that in this reality we don't understand why one reality outcome is more likely than another (Many Worlds).

Another you might ask is how these two basic interpretations* might be tested or falsified or decided between since they are both compatible with the observed test results. I will say one way in which MW is falsifiable: if a split between the quantum and macro world IS ever found, then that would falsify MW, because MW's core assumption is that all of reality is really QM. Such a split would show that, no, there actually ARE two discrete levels of reality. So far, though, we've manage to put over 2000 particles in superposition simultaneously; so, if there is some "split," it isn't to the 2000th (or less) particle. As for the other questions, well, I'll save that for if you want to hear my "rhetoric" ("rhetoric" in this case being my underlying philosophical beliefs, epistemic rationality using Bayes' Theorem, that lead me to vastly prefer MW).

I should also stress that, according to your post, there is really only the "known" of QM (known either via experiments, like the double-slit, or via mathematical proofs like Bell's Theorem) and the various "interpretations" (speculations) on what that "known" means. There's really not anything in the middle unless you privilege some sort of philosophical standpoint. Two examples: if you privilege Occam's Razor, or the idea that the simplest explanations are the most likely, then MW is demonstrably simpler than CI because CI is the interpretation "adding" things (namely, the collapse) to explain what we observe. MW is just assuming what we see is real and that that's what all reality is like. If you privilege human intuitions, then CI probably seems more palatable than the notion that there are other worlds and versions of us in those worlds that we can't contact or confirm in any tangible ways. It's easier to imagine a split between a singular world where there is a "micro" and "macro" level.

*I say two "basic" interpretations because there are certainly others, but most interpretations are off-shoots of either the "collapse" or "decoherence" principles. EG, "Many histories" and "Many Minds" are offshoots of the "decoherence" principles. So, the first step is to think about whether you find decoherence or collapse more plausible, given what both assume and what problems they create.

Hopefully all of this was clear/lucid. If you have any specific questions, just let me know.

The mathematics is just a tool to describe what is real. It is a map of the territory, not the territory. (I'm trying to quote something I think Yudkowsky wrote.)

Yudkowsky appropriated the map/territory distinction from Alfred Korzybiski. Anyway, you're absolutely right. However, for mathematics to be a map there has to be a territory that it's modeling. If you're assuming the wavefunction is nonreal then there's no territory to model!

So, overall the many worlds there is determinism and localism, but within each of the those worlds QM must work as it does in our world. Is that what MW claims? Since we can't ever get to a perspective where we can see this determinism and local behavior, what good is it?

We can't "see" localism and determinism because we ourselves are part of the system we're observing. The mistake MW proponents think Copenhagen is making is removing the observer from the process of the observation. MW says that the observation is an instance of two quantum systems (observer and observer) becoming entangled and ceasing to be in a state of superposition. It's this entanglement that makes MW local (observation can only happen locally); it's the fact that all outcomes are realized that makes it deterministic.

Since I don't think MW is true, he would be wrong about that.

What I meant is that you can't point to anything in his articles that are factually incorrect. Yudkowsky is making the case for why he feels MW is overwhelmingly more likely than CI. If you want to object to Yudkowsky's sequence, you need to object to one of his arguments that lead to the conclusion "MW is most likely true" rather than just objecting to the conclusion.

[Calidore]

Morpheus: Thanks very much for taking the time to write all that out. I'll no doubt have plenty of questions once I've been able to read it in depth. I do tend to believe in going with Occam's Razor when possible, but I don't agree that necessarily favors MW. MW may have the simplest path, but leading to the wackier conclusion. CI has the simpler conclusion, but it requires taking many liberties to get there. That's why I suspect the truth will turn out to be a "C" choice, whatever form that may take.

You also state further above that if our knowledge of QM was incomplete, our predictions wouldn't be so accurate; I submit (granted, much more uninformedly) that holes in knowledge don't always look like holes until you step on them and start falling. How many times have the space guys been completely surprised by probe findings and had to scrap a whole bunch of ideas that had made perfect sense until then?

I think I have also come to understand your much earlier 80/20 loaded coin example in the context of the weather: When the forecast has a 50% chance of rain, that 50% chance is an illusion; it's actually 80% if you go outside and 20% if you stay in.

[MorpheusSandman]

You're very welcome, Calidore.

I don't agree that necessarily favors MW. MW may have the simplest path, but leading to the wackier conclusion. CI has the simpler conclusion, but it requires taking many liberties to get there.

That MW is simpler than CI is one the "facts," not speculations. I mean if you literally wrote out those interpretations as mathematical code, MW would be much simpler because it's not "adding" anything to the formulas that model QM, it's just treating them as real. CI is treating them as non-real, adding the collapse, and then having to face the various problems this creates.

That a simple starting point leads to a more complex conclusion shouldn't surprise us as that's precisely what we see in nature; simple, fundamental things aggregating to more complex forms. This happens in everything from evolutionary biology to societies to matter itself (quanta aggregate to particles to atoms to molecules). This is why I stressed that you have to judge MW based on the assumptions being made, not where it ends up.

You also state further above that if our knowledge of QM was incomplete, our predictions wouldn't be so accurate; I submit (granted, much more uninformedly) that holes in knowledge don't always look like holes until you step on them and start falling. How many times have the space guys been completely surprised by probe findings and had to scrap a whole bunch of ideas that had made perfect sense until then?

It's not just an issue of accuracy (though it is scary accurate: Feynman once famously said that it's as accurate as predicting the the distance of the width of North America to within the span of a human hair, and that's not really an exaggeration), it's an issue that, as Bell's Theorem proved, no hidden variables could account for all the predictions of QM while maintaining the principle of locality under the assumption of collapse.

I think I have also come to understand your much earlier 80/20 loaded coin example in the context of the weather: When the forecast has a 50% chance of rain, that 50% chance is an illusion; it's actually 80% if you go outside and 20% if you stay in.

No, that's still wrong. The important thing to understand is that reality has no probabilities. The probability any thing will happen is always 100%/0%. When we humans assign probabilities, those probabilities express a combination of how much we know and don't know about the subject in question. In the coin flip, we know the coin has two sides; we don't know (or can't calculate) the various physical forces that will determine what side it lands on. In the weather example, we know what meteorological situations are most likely to produce rain; but those systems are so complex that we can not perfectly track every development to decide if/when it will, indeed, rain.

Perhaps a better example than the coin or the weather is a shuffled deck of cards. Once you shuffle a deck, the top card is what it is. If the top card is the queen of hearts, the probability it's the queen of hearts is 100%. However, if we were to assign a probability to the top card being the queen of hearts (without having looked at it), we'd say the probability is 1/52 or 1.9%. That probability expresses both what we know about the situation (there are 52 cards), and what we don't know about the situation (what the result of the shuffle was).

[Calidore]

No, that's still wrong.

That was actually meant as a joke.

[MorpheusSandman]

That was actually meant as a joke.

[Nick Capozzoli]

The double slit experiment is really fascinating (electrons as waves or particles?), but there is another famous physical paradox from the field of thermodynamics that is worth considering. That is "Maxwell's Demon." Basically this involves a hypothetical experiment in which we start with two chambers containing gaseous molecules at the same temperature. There is a sort of "gate" that can be opened or closed by a miniscule "demon" in one of the chambers who is able to observe molecules moving around in his chamber. In particular, this demon is able to determine the speed and direction of a molecule approaching the gate, so as to allow, if the demon wishes, the molecule to move to the other chamber by opening the gate to allow its passage to the other side. What Maxwell wanted to know was whether or not the Demon could effect a temperature change between the chambers, for example by allowing faster moving particles to pass through the gate and thus causing an increase in the temperature of the chamber on the other side of the gate.

This is a profound question, and every bit as profound as the question of interpreting the wave versus particle behavior of electrons in the double-slit experiment. Thermodynamic "laws" are essentially "empiric" laws, that we believe are "empirically" true on a macroscopic level. Thermodynamics, like quantum mechanics, depends on "statistical" mathematical analysis to explain the behavior of matter.

Maxwell's Demon involves a "thought experiment." The main conundrum is that if we allow the Demon to be able to control the gate in such a way that he can allow fast-moving particles to enter the "other" chamber, the Demon will violate the Second Law of Thermodynanics.

There has been much discussion about this Maxwell's Demon, and mainly why this Thought Experiment doesn't lead to any physical problems or logical uncertainties...

I think that a way of dealing with the problems of the putative "Demon" involves an appreciation of the fact that the Demon, in order to be able to do what Maxwell wanted him to be able to do, would need to have a certain degree of "molecular organization. The 'organiational complexity" of the demon would have to be figured into calculations of the ability of the demon to "violate' the Second Law.

[YesNo]

I guess the demon would be providing an energy source to do the work. http://en.wikipedia.org/wiki/Maxwell's_demon

It does make me wonder how if everything runs down, gets entropy, what is the physical law that allows something to lose entropy overall in a closed system, not just transfer less entropy from one part to the other? Chance isn't going to work here. Chance is what increases entropy.

Getting back to quantum stuff, such as electrons, viewed metaphorically as particles, they don't ever seem to run down. They have to keep moving like a perpetual motion machine. If they did take a break, then we would be able to know their precise position, x, and precise momentum 0, which would violate QM. The particle metaphor breaks down.

This is why I stressed that you have to judge MW based on the assumptions being made, not where it ends up.

The assumptions are tested, "judged", by the logical results, that is, "where it ends up".

If one ends up with a logical contradiction, and the chain of reasoning is correct, then one or more of the assumptions is wrong. If one ends up with a conclusion that violates experimental results then one or more of the assumptions, or some reasoning along the way, has to be rejected.

The end results provide the basis for falsifiable tests.

In many worlds nothing seems falsifiable. Do an experiment and in one set of worlds you get one answer. In another set of worlds you get another answer. Why bother?