Figs 1 and 2 show the setup. Here's the conclusion:
> "This gives a clear proof of the compatibility of constructor-based irreversibility with unitary quantum theory, providing a frame for the emergence
of thermodynamical irreversibility in quantum mechanics."
It seems like an extraordinary claim, but is very interesting. The theoretical framework (constructors) is a bit confusing. Hopefully this will be replicated.
I recall someone talking about something rather similar once, related to so-called chaos theory (now more generally called 'sensitive dependence on initial conditions'), which describes how deterministic systems can give rise mostly unpredictable phenomena (strange attractors etc.), in which indeterminate quantum states at the atomic level are the ultimate cause of the phenomenon.
dreamcompiler 672 days ago [-]
Chaos theory is orthogonal to quantum theory. Neither depends on the existence (or nonexistence) of the other.
Chaos is unpredictable but for a different reason than quantum systems: Chaotic systems are unpredictable because we can never measure the real world perfectly (even ignoring quantum weirdness) and that lack of precision in measurement coupled with nonlinearity means our measurement errors amplify quickly. Nevertheless, chaos is theoretically deterministic. It's just not practically deterministic (except in the coarse-grained statistical sense of saying "the system's position in phase space is somewhere on a strange attractor path bounded by X.")
Quantum systems are not even deterministic in theory, let alone practice. And ironically quantum systems are fundamentally linear. So the spookier theory in a sense is based on more tractable math.
tsimionescu 671 days ago [-]
> Quantum systems are not even deterministic in theory, let alone practice. And ironically quantum systems are fundamentally linear.
It's important to note that the parts of quantum mechanics that are linear are also fully deterministic - the Schrodinger equation predicts the evolution of the wave function of a system with exact precision (as do the more complex equations of QFT).
The part that introduces non-determinism is the non-linear adjustment of the Schrodinger equation that you have to perform after a measurement - the Born rule, also called wave-function collapse. Here the complex-valued wave function of a system randomly takes a single real value at the moment of measurement. It does go back to its deterministic linear evolution afterwards though, but you have to perform this non-linear adjustment.
This non-linear adjustment can be interpreted in many ways, but without it QM doesn't agree with even the most basic experiments, so one way or another it is a vital part of the theory, rendering it non-linear overall.
dreamcompiler 670 days ago [-]
Well-said. I was a bit sloppy in my terminology because the math of quantum computing is mostly linear (albeit complex) but I was thinking of the Schrodinger part.
When qbits have to be read, we convert them from complex to real and that step is both nonlinear and nondeterministic.
672 days ago [-]
mjburgess 672 days ago [-]
Chaos is only deterministic with infinite precision. Since the universe isnt the kinds of place where position-momentum etc. have infinite precision, chaos isnt deterministic.
Chaotic orbits of whole moons are non-deterministic on quite short time horizons.
leephillips 672 days ago [-]
Since all these phenomena of chaos are inherent in the differential equations, how can they possibly have anything to do with quantum states? The sensitive dependence on initial conditions is a mathematical property of the description of the chaotic systems. It’s a closed mathematical system that makes no reference to any physical reality.
photochemsyn 672 days ago [-]
Here's the quote, it was in some talk he gave:
The chaos can act as a magnifier of quantum fluctuations so that they can produce sizable effects in the world around us. But we know that that can happen often. - Murray Gell-Mann
I guess the idea is that if you have a collection of particles in a system that displays sensitive initial dependence, then any quantum jiggling gets rapidly amplified.
[edit] As you note this doesn't explain the chaos phenomena, because a stable dynamic system would squelch any quantum weirdness rather than amplify it.
clairity 672 days ago [-]
yes, chaos theory, including strange attractors, is still deterministic. it acts at a level where the tiny probabalistic effects of a multitude of tiny particles cancel out in ways to miraculously make the larger subjects of interest predictable.
it's the same phenomenon underlying the wisdom of crowds (and the pricing function in fair markets): each individual can be wildly incorrect in an observation, but averaging many (independent) observations together reveals stable insight.
grungegun 672 days ago [-]
The differential equations which give rise to physical chaotic systems operate on quantum states interacting with each other at the atomic level, so there has to be a link of some kind.
leephillips 672 days ago [-]
I don’t understand your comment. Differential equations don’t give rise to physical systems.
Some nonlinear DEs exhibit chaos. That’s a purely mathematical property. Whether any particular DE is a useful model of a particular physical system is a matter for the imagination, and either backed up or refuted by experiment.
grungegun 672 days ago [-]
> Differential equations don’t give rise to physical systems.
Sure, but differential equations describe physical systems, and there is a canonical way to derive a quantum differential equation from a classical one by quantifying the classical Lagrangian using the Path Integral formulation. Giving a fairly natural distinction between the types of equations
> Whether any particular DE is a useful model of a particular physical system is a matter for the imagination, and either backed up or refuted by experiment.
This doesn't make sense to me. The Navier-Stokes equations are known to describe the classical behavior of water and are experimentally confirmed to predict things like trajectory. Their effectiveness has nothing to do with my imagination. If I write x=x' for the position vector of atoms in a fluid that will completely fail to describe anything physical.
leephillips 672 days ago [-]
The NS equations are a good description of approximately Newtonian fluids like water within certain regimes (nonrelativistic velocities, larger than atomic scales, far above 0°K, etc.). “Imagination” was not the best choice of word on my part. I meant that DEs are mathematical objects; their connection to physical systems is made by the scientist, not inherent in themselves. Whether the scientist guessed right is determined by experiment. The first people to suggest that the NS equations were a good description of Newtonian fluids had a model for fluid behavior in their imaginations. We know it was a good model because of experiment. But even if the NS equations described nothing in nature, their solutions, chaotic and otherwise, would have whatever properties they have.
Note that there is no Largrangian for the NS equations, by the way.
grungegun 672 days ago [-]
Yes, the quantumness of a differential equation is not a property of the differential equation, but a statement about one possible taxonomy of differential equation. Then, whether quantum-type diff eq's have unique properties pertaining to chaos conditioned on our knowledge of them being labelled 'quantum' is an interesting mathematical question.
> Note that there is no Lagrangian for the NS equations, by the way.
I don't know much about fluid dynamics, but I was under the impression that Bennett derives the Lagrangian form in the book Lagrangian Fluid Dynamics
leephillips 672 days ago [-]
There is a clash of terminology: the Lagrangian formulation of fluid dynamics follows the path of fluid particles, in contrast to the Eulerian form, which observes the fluid passing by a fixed coordinate system. In general dissipative systems don’t have time-independent Lagrangians.
grungegun 672 days ago [-]
yep, that sounds right to me. also I see you're the author of the Noether article from Ars Technica. I enjoyed the read.
It looks like in the results section eqs (8)-(13), they give the main thrust of their idea. To demonstrate that not all machines are reversible they have to take the limit of the number of qubits of their machine as n goes to infinity, so their result is still in the limit, same as the typical laws of thermodynamics.
So this is primarily an interesting mathematical restatement at the quantum level. I don't know how useful this is, not having much knowledge about existing quantum information theory/thermodynamics.
lowwave 672 days ago [-]
> sensitive dependence on initial conditions
isn't that the wolfram thing? Simple patterns create complex unpredictable patterns? And our universe are just made up a few very simple rules and in combination it created this complex universe. Yeah most ideas has being around for a long time, it is just now we have the technology to verify some of these claims.
derbOac 672 days ago [-]
It's hard to tell from the Quanta article, but without meaning to sound dismissive, a lot of this seemed like reexpressing some of the same ideas in different ways. The constructor and resource theories seemed very similar to me, for instance.
I was also extremely puzzled by the dismissal of probabilistic or statistical models as being fuzzy or something. Probability has always been the domain of possibility -- this is basically the Jaynesian-style account of probability. It's always seemed to me to be the case that at some level statistics is about selective loss of information -- that is, as you go from descriptions at the particle level to group-level statistical accounts, you summarize (in a coding sense) the ensemble of particles' behavior to shed information required to maintain complete information about the ensemble. To make an analogy, if you wanted to know people's heights, at some point storing information about 1 million person's heights would become burdensome, so you'd replace it with a single number (e.g., the mean) instead.
That's not to say what's being discussed in the article isn't interesting, but it just seems like a different level of analysis or something? There's also different ways of representing amount of information in the quantum realm vs classical realm, so there's that, but it's also the case that a chamber of gas will generally involve a lot of molecules in many classical cases.
I also had the sense that there was probably something more fundamental that you could invoke that would integrate a lot of this, at the level of logic or computational theory or something. But maybe at this level the difference is sort of moot.
thingification 672 days ago [-]
> I was also extremely puzzled by the dismissal of probabilistic or statistical models as being fuzzy or something. Probability has always been the domain of possibility
Perhaps you're just saying that probability is useful in some practical ways? Deutsch agrees (see video below). But if that's not all you're saying:
If the "domain of possibility" (which I interpret to refer to quantum mechanics and thermodynamics) can be described without the whole concept of probability, why not eliminate it? Why eliminate it, though? As Deutsch says "We don't need a reason: Physics likes to do without things"? [1]
You may say constructor theory adds new fundamental concepts. I don't know constructor theory so I don't know whether that's true. In cany case, what's not to like about probability concepts? In what way are they "fuzzy or something", as you put it? Deutsch points out here [2] (few minutes later in the same talk as the other link) a sense in which probability statements don't describe physical facts: for example "drawing to an inside straight is very risky" is logically consistent with any sequence of cards in a poker game. I think that's the "fuzziness" they want to eliminate from Physics. He then goes on to deal with some objections to that idea such as "in the long run, probability statements describe physical facts".
If the game of Physics is describing physical reality, and we don't need that specific kind of fuzziness, success would seem to me an advance in Physics.
Obviously probability is great and all, but the second law really really holds. Honestly, you could tell me you made energy out of nothing and I wouldn't rule it out unless you could do work with it. But like, why should every physical law we look at have states distributed in such a fashion? Especially with information brought in, and worse, quantum information? You're right that this is mostly reshuffling but it's still neat, and helps give an intuition for why we'd expect it to work out like this
dgellow 672 days ago [-]
I know very little about actual physics, but generally speaking the more grandiose headlines sound the less I pay attention…
Could someone who knows the topic tell us (and by “us” I mean layman like me) if there is something to keep from this article? It brings so much context that after reading it I still have no idea what this “rewrite of fundamental law” is about.
boole1854 672 days ago [-]
My summary:
There has always been something unsatisfactory about the second law of thermodynamics. Unlike many other laws in physics, which precisely state both what kinds of structures exist in reality and how they behave, the second law has a built-in lack of precision. Specifically, there are multiple statements of what the second law says, but they all suffer from referencing some concept which is inherently fuzzy at some scale. For example, they may refer to "heat" as part of the law, but the physical definition of "heat" is fuzzy at the microscale -- it only has a clear meaning for "large systems".
However, using a new physio-mathematical formalism called "constructor theory" the researchers have now found a way to state the second law of thermodynamics which is precise at all scales.
whatshisface 672 days ago [-]
Constructor theory is philosophy, not physics or math. It is concerned mainly with how to state things, not how to discover things. There is nothing wrong with metaphysics (it is made necessary by the nature of human language and cognition) but I feel like people should know which theories are what kind.
boole1854 672 days ago [-]
I wonder if you could explain your critique in more detail. One of the constructor theory papers under discussion in this thread describes a physical experiment that was done [1]. In what sense is this not physics?
The paper contains an experiment analyzed according to the pre-existing theoretical framework for that kind of experiment, and it also contains some phrases phrased in the language of constructor theory. You could take the constructor theory language out and have the same paper.
In the scientific sense, it is not enough to show that your system agrees with experiments to say it is true, you also have to show that the alternatives do not agree with experiments. Constructor theory is metaphysical because you'd get the same analysis whether you used it, ad-hoc intuitive thinking (in practice, people do not need instruction on universal principles like cause and effect), or some other alternative. All of the non-intuitive weight in that paper is borne by the QM stuff that exists independently of constructor theory, and in contrast, if you changed that the answers would start coming out wrong.
I am not proposing that this captures the essence of positivism, but one test that you can apply to check new ideas without going through the trouble of understanding them is to see whether they introduce any new conclusions that could not have been obtained under previous frameworks. If not, they may be metaphysical re-arrangements of existing ideas.
boole1854 672 days ago [-]
It seems to me that a theory which uses precise, hard-to-vary explanations to make falsifiable statements about observables is in fact a scientific theory.
One of the points of the paper is that constructor theory makes such statements in a way that is compatible with other known laws. This is demonstrated via the experiment. So I'm afraid I do not understand your remark that "you could take the constructor theory language out and have the same paper". Taking the constructor theory language out would undermine a major point the authors are trying to make in the paper.
whatshisface 672 days ago [-]
>One of the points of the paper is that constructor theory makes such statements in a way that is compatible with other known laws.
Yeah, but known laws are already compatible with themselves. You might make them more comprehensible by establishing a conceptual meta-framework that offers a consistent way to phrase statements about them, but that is clearly a philosophical task.
bordercases 672 days ago [-]
The fact it is philosophical does not make it irrelevant to physics. As such to point out the distinction here does little work. I.e. What QM interpenetration do you support and why are the others unendorseable?
wisty 672 days ago [-]
You could kind of say a similar thing about some of Maxwell's and Einstein's work - restating Faraday or Lorentz in a more elegant way is not great achievement, but if it leads to a clue where a new rule can be found then it might be very useful.
But I'll admit I don't understand constructor theory, and maybe it can hide any interesting discrepancies by overfitting.
whatshisface 672 days ago [-]
Maxwell added a new term to what are now called Maxwell's equations, called the "displacement current," completing them and allowing waves to exist in the theory. Although the symmetries in Maxwell's equations that Lorentz observed were known before Einstein published his paper on special relativity, Einstein was the first to make the bold suggestion of applying them to all matter, not just electrodynamics. In fact, during the period when people believed in the ether, those symmetries in Maxwell's equations were considered flaws in the theory that would be corrected when the properties of the ether became known.
guerrilla 670 days ago [-]
I suspected that when I was reading the article but I'm not convinced. What is your background in this area?
namero999 672 days ago [-]
A tiny adjustment to an otherwise ok summary: physics state how nature behaves, and not what structures exist in and of themselves. All structures posed by physics are metaphorical proxies stemming from our perception, not real structures.
boole1854 672 days ago [-]
I suppose I should say that I am using the understanding of "physical explanations" promoted by David Deutsch, among others. (Deutsch is the creator of constructor theory which is the subject of this thread):
"A scientific explanation is a statement of what is there in reality, and how it behaves and how that accounts for the explicanda." (emphasis added) [1]
Deutsch argues strongly against the view of physics which says that physical explanations are just about predictions of behaviors and not about what is really out there. See also his 1997 book The Fabric of Reality.
"A scientific explanation is a statement of what is there in reality"
Sure, you can say stuff about "what is there in reality", but I don't see any way one can verify such statements.
Everything humans have access to is mediated through their perceptions. Everything that we sense "outside" of us is to us a perception.
This includes the results of all the experiments ever done, all our interaction with tools and scientific instruments, and all our observations. They're all perceptions.
Our thoughts and feelings are also perceptions. The results of our deductions and inductions are perceptions.
Humans seem to be hermetically sealed from the "outside world" (if there is one) by our perceptions, and I don't see any way out of that... not through science, not through philosophy, not through religion, not through anything.
boole1854 672 days ago [-]
You may find the following journal article relevant and interesting:
In short, late in the 19th century and early 20th century, the physics philosophy of Ernst Mach held great sway. Mach's viewpoint was the very viewpoint that you describe: the only things we know about are our own perceptions and proper formulations of physical theories consists at the fundamental level of descriptions of these things.
Mach had a significant impact on many physicists' understanding of what they were doing and also had an impact on the development of logical positivism in the field of philosophy.
Einstein accepted Mach's philosophy initially. But gradually he came to see it as untenable given his own work in physics. The article linked to above describes how and why his viewpoint changed over time.
pmoriarty 672 days ago [-]
Thank you for a very enjoyable paper. It was interesting, as you said.
However, if I gave you the impression that I was a Machian, this is a mistake.
According to this paper, Mach equated reality with sensory experience (ie. the experiences afforded us through the five senses: seeing, hearing, touch, smell, and taste.. and perhaps some others like proprioception), and that speculation about a world "outside" and independent of such sensations was unwarranted.
This is not my position, as I include in "perception", the perception of thoughts, theories, conclusions, intuitions, feeling and everything else considered to be "internal" -- which includes everything Einstein or any other physicist could ever base their own theories on.
Also, I don't object to speculation as to an "external", "objective" world, as Mach does. Science can come up with all sorts of interesting and possibly useful things through such speculation. I'm happy to read of such conjectures, and if some people want to engage in it more power to them.
However, do I not see any way we could ever verify whether such speculation right or wrong (or, to put it another way, whether such speculation corresponded to the way the world really was), because neither mathematical simplicity or the elegant fitting in to other, larger theories (as Einstein seems to have favored) means that the speculation has anything to do with the "external" world as it is (or even that there is an external world).
Since we seem to be limited to our own experience, the belief that an "external" world exists and that we have any (direct or indirect) way to access it seems to be nothing but faith.
rootw0rm 672 days ago [-]
This is why I love HN. Thanks for contributing, this has been a very interesting discussion.
namero999 672 days ago [-]
I have read that book, but I personally disagree with such a premise. I hold that in the moment a scientist or otherwise poses a statement in the form of "X exists in and of itself", they are doing philosophy, not science.
The scientific method is about predictions and testability of such predictions, therefore it implies behavior. Nothing more is needed for a scientific theory. To test for existence within the scientific method would require to test for non existence too, a clear paradox, which is the reason why ontological statements ("X exists") are meta-physical in nature.
boole1854 672 days ago [-]
> I hold that in the moment a scientist or otherwise poses a statement in the form of "X exists in and of itself", they are doing philosophy, not science.
How far do you take that stance?
If a biologist says a particular bacterium exists, are they doing philosophy, not science?
Edited to add another interesting case: creation "science" organizations like Answers in Genesis say that the theory of evolution is a philosophy not science since it describes non-observables in the past (events in the past are inherently not observable now). When a paleontologist says a particular animal once existed in the past, would you say they are they doing philosophy not science?
namero999 672 days ago [-]
I take it all the way _down_ to the reduction basis, which is the only place where all of this matters.
When a biologist says that a bacterium exists, they are making a descriptive statement, not an ontological one. To say that a bacterium exists is just a convenient proxy to refer to a specific aggregate as a whole. A bacterium is always reducible to its constituents, therefore has no ontological status nor existence in and of itself: in other words, if the molecules which the bacterium is made of would relinquish their structural integrity, the bacterium would cease to "exist" too.
Therefore I would agree with you that it's totally legitimate for a scientist to claim that a bacterium exists, in so far as we understand this claim as a descriptive proxy. A shorthand, essentially, to cluster together specific configurations of the underlying, supposedly more real, "stuff". And this is needed, if we want to make any progress. We have to communicate, effectively and succinctly. Science is precisely about _descriptions_ and predictions of behavior. Same goes with the paleontologist example.
It's a completely different matter instead to claim ontological status, that is, to say that something exists in and of itself, irreducible to something else, as a building block of the universe. This is a philosophical stance per-force, since the scientific method cannot follow us there. It's meta-physics by definition, and there are different rules at play there. Not Popper's falsifiability, but empirical adequacy, internal consistency, explanatory power, etc.
So when scientists claim that the ontological primitive is, e.g., the quantum field, or "energy" or anything really, they are stepping soundly and firmly into philosophy. Which is totally fine to do. We just need to be aware of this, and start to apply different rules to evaluate the validity of the claims.
Edits are typos.
boole1854 672 days ago [-]
I see. Now I understand that when earlier you were adding the "in and of itself" qualifier to "exists", you meant the same thing by what I would call "exists at an ontologically fundamental level". I agree with that.
However, note that in my original post I didn't claim that physical laws describe what exists at an ontologically fundamental level, just what exists in reality. And by that I mean "exists" in the non-subjective way that, say, hydrogen atoms exist. I wouldn't mean to imply that hydrogen atoms are irreducible.
Does Deutsch and his research group make claims to know what is ontologically irreducible? Not as far as I know, but I am open to correction!
dgellow 672 days ago [-]
Thank you very much!
alpple 672 days ago [-]
"I know very little about actual physics, but generally speaking the more grandiose headlines sound the less I pay attention…"
You've managed to find the words that I couldn't find to explain how I feel. Thank you.
ThePhysicist 672 days ago [-]
The article is of course oversimplifying, but if I understand correctly the irreversibility in the qubit system comes from the fact that they assume you don't know the states of the qubit chain that transforms the first qubit from B_1 to B_2. If you'd know that the state would be pure and it would be easy to reverse the entanglement, it's just that if you look at the state of the first qubit alone that it's mixed.
I actually envisioned a similar thought experiment that could be done with a very complex quantum computer (say millions of qubits) to test whether the universe is actually keeping tabs of all these states or if it gives up at some point and produces irreversibility. But I guess if you'd formulate it like that the academic community wouldn't take you very seriously anymore. I'd wager Deutsch's actual motivation might be along this line, as he seemed very interested to find a way to test the many worlds interpretation against other interpretations of quantum mechanics.
boole1854 672 days ago [-]
Deutsch's position is actually that the many worlds interpretation has already been tested and has passed the tests so far, whereas rival theories have been refuted.
To summarize the argument that the blog post makes:
The multiverse interpretation accounts for cases where the probabilities of each distinct observable outcome are equal according to the Born rule. But it cannot account for cases where the Born rule says there are non-equal probabilities for the possible outcomes.
This is a fairly common problem which critics of the multiverse interpretation raise. See [1] for another example where a critic raises that problem in more detail (with examples).
Deutsch and others have responded. In short, the reply is that the mathematics of the Schrodinger equation describes a density of universes in the continuum which is the multiverse, not a finite set of distinct, interacting single universes. This undercuts the critics' argument. See [2] for an example of this kind of response.
No, it doesn't, for all of the reasons given in the blog post and the subsequent discussion. The Born rule is an observable fact of nature, and you cannot derive it from the Schroedinger equation, Duetsch and Wallace's claims to the contrary notwithstanding.
boole1854 672 days ago [-]
Maybe you can help me understand your argument better. If the Schrodinger equation describes a density of universes, doesn't that undercut the argument in the blog post?
Edited to add my understanding: if the cardinality of the multiverse is uncountable, one has to talk about densities of branches. In that scheme, the problems which branch counting introduce do not arise.
lisper 672 days ago [-]
> If the Schrodinger equation describes a density of universes,
It doesn't (see below) but even if it did...
> doesn't that undercut the argument in the blog post?
No. It's really very simple: the SE is linear. The BR is not. You can't derive a non-linear result from a linear equation.
The only thing that is even a tiny bit complicated about this is figuring out where the Deutsch-Wallace argument fails, because Deutsch and Wallace are not idiots and they hid their questionable assumption very well.
So... why does the SE not describe a "density of universes"? It's because the SE does not describe universes at all! Deutsch and Wallace specifically acknowledge this, which makes their argument all that much more untenable. The SE describes the behavior of a function called "the wave function". The wave function is not a description of anything physical. The domain of the wave function is not physical space, it is configuration space. This is the reason entanglement seems weird to us, because we live in physical space and the wave function does not.
In order to get anything physical out of the wave function -- particles, universes, anything resembling our classical reality -- you have to add some additional assumption, and any additional assumption that you add will be in some sense isomorphic to the Born rule. There is no way around this. Trying to derive the BR from the SE is like trying to prove Euclid's fifth postulate.
hither_shores 672 days ago [-]
> The wave function is not a description of anything physical. The domain of the wave function is not physical space, it is configuration space. This is the reason entanglement seems weird to us, because we live in physical space and the wave function does not.
This is circular: you're assuming that realism about the wave function is incorrect to argue against the (tragically misnamed) many-worlds interpretation, but the only substantive claim the MWI makes is that realism about the wave function is true.
lisper 672 days ago [-]
> you're assuming that realism about the wave function is incorrect
No, I'm not assuming it. I simply observe that the SR by itself does not describe the outcomes of experiments. And I'm not even talking here about quantum experiments. The SR cannot by itself even account for the fact that we live in a three-dimensional space or that there is an arrow of time. And it certainly cannot account for the Born rule, which is also an experimentally observed fact.
> the only substantive claim the MWI makes is that realism about the wave function is true.
Yes, I know. The problem with this is that it is violently at odds with observations that every human makes every moment of their lives, and it does not account for the Born rule.
Other than that there's no problem with it.
hither_shores 671 days ago [-]
> I simply observe that the SR by itself does not describe the outcomes of experiments.
No formalism describes the outcome of experiments by itself. There always needs to be an interpretation attached, and our choice of interpretation is always underdetermined by observation.
> The problem with this is that it is violently at odds with observations that every human makes every moment of their lives
Like all other interpretations of all other physical theories, MWI isn't and can't be at odds with our observations alone, because our observations alone aren't enough to imply any particular relationship between observations in general. And like all other interpretations, it is and must be at odds with competing interpretations, and in particular with the folk-physics model we're naturally inclined to use. But this means very little, because folk-physics is wrong, and this far from the paleolithic savanna, it's wrong in ways that matter. Intuition is evidence, but it's not particularly strong evidence, and it's certainly not proof.
> and it does not account for the Born rule.
Sure, it still needs a separate account of subjective probability. But so does everything else: it's just a little harder to notice in the classical regime, since the relevant probabilities are always either 1 or 0.
lisper 671 days ago [-]
> No formalism describes the outcome of experiments by itself.
That is so wrong as to border on the ridiculous. The quantum mechanics formalism, i.e. the SE plus the BR together describe the outcomes of experiments. No interpretation is required. You just turn the crank on the math and predictions pop out. There is a reason that "shut up and calculate" is a thing.
The only reason people talk about interpretations at all is that the structure of the math is violently at odds with our intuitions and subjective experience, and this leaves some people feeling queasy and wanting an explanation. There's nothing wrong with that. But it has absolutely nothing to do with the power of the formalism to make predictions, which is completely independent of any interpretation. (And this is true of classical mechanics as well.)
zarzavat 672 days ago [-]
Ultimately the MWI is saying “believe the math - configuration space is physical reality”. So you are on solid ground when you give arguments that are mathematical, however the metaphysical stuff about the self is basically talking past the MWI, which specifically does not acknowledge that as valid form of argument. I also agree with the post above that arguments about “branch counting” are bunk because you cannot count in a continuous space (see: banach-tarski and related paradoxes).
lisper 672 days ago [-]
No, the MWI says to believe half of the math and to ignore the other half. The problem with that is that the second half is what allows the first half to make experimental predictions. Without the second half, there is no reason to believe the first half.
spekcular 672 days ago [-]
Your reference [2] is somewhat suspect. As far as I know, it is not generally accepted in the physics community that the Born rule can be derived directly from the Schrodinger equation. In fact, it seems generally accepted that such a derivation does not currently exist, and MWI has to stick it in as an additional postulate in some form (or otherwise give an auxiliary argument for it, none of which seem generally accepted either).
I don't have time at the moment to review the paper in detail to locate any potential flaws, but I wanted to flag this for others. I think there is substantial reason to be skeptical.
boole1854 672 days ago [-]
I agree that Tipler is far off the deep end on on some topics. But you should check out that one paper I linked nevertheless. PNAS is not a junk journal.
I also agree there is no consensus that the problem of getting Born's rule from the Schrodinger equation has been solved to everyone's satisfaction.
spekcular 672 days ago [-]
PNAS publishes a lot of junk. If you're connected to the Academy you can get papers through with very light review.
In fact, one can even seen this happened in the Tipler paper: "This Direct Submission article had a prearranged editor." [0]
Edit: I read the paper. It is borderline crank work. There are decent arguments for the Born rule in MWI but this is not one of them. (I like David Wallace's book, for example.)
boole1854 672 days ago [-]
Care to elucidate some critiques of the paper?
spekcular 672 days ago [-]
Care to outline a precise argument for me to respond to? I'm honestly not sure one can even be extracted from pages 3 through 5, and I don't want to attack a strawman.
One basic problem is that using symmetry arguments to justify the assignment of subjective probabilities (cf. page 4, right column) is going to fail when there isn't a symmetry available, for example for irrational probabilities.
A deeper issue is the implicit assumption that one can coarse-grain into discrete "effectively distinguishable universes" for the purpose of doing these counting arguments. There are significant conceptual difficulties doing this, and this claim is much stronger than the bare MWI. Wallace discusses this at length; see for instance the citations in the blog post you responded to.
ahazred8ta 672 days ago [-]
Perhaps it occupies the gray area between "wrong" and "not even...". O:-)
RandomWorker 672 days ago [-]
Laws in physics are written based on testable hypotheses and measurements to confirm. This is a constant process as the hypotheses improve and the measurements improve laws will change. This is the way.
mc4ndr3 672 days ago [-]
Both conventional probability and quantum mechanics are worrisome: On a long enough timeline, the universe will (repeatedly) reset itself to precisely the original state at the time of the big bang. It is difficult to avoid a universe where time is not cyclical. Especially if space turns out to be finite. Or if indeed parallel worlds do exist. Or if a big crunch happens.
edgyquant 672 days ago [-]
Sigh. There is no link to a peer reviewed paper in there, just describing entropy and a few scientists working on describing thermodynamics via quantum theory.
Then this paper applied constructor theory to form a constructor theory of information:
Deutsch David and Marletto Chiara 2015 Constructor theory of information Proc. R. Soc. A.4712014054020140540
(Download: https://royalsocietypublishing.org/doi/10.1098/rspa.2014.054...)
Finally, this paper applied the constructor theories of information and probability to thermodynamics:
https://arxiv.org/pdf/1608.02625.pdf
(not yet published in a journal, to my knowledge)
mindcrime 672 days ago [-]
It appears that the first occurrence of the the name of each of the researchers mentioned in the article links through to their respective home-page. At a quick glance, most of them appear to publish lists of their publications, some with direct links to the actual papers. Of course it might be hard to tease out which papers they've contributed to that pertain to this specific topic, but it would probably be worth the effort for someone who was really interested.
It would be nice, but I think this is a survey of an evolving field and there probably isn't any one pivotal paper. Plugging some key names (Marletto, Halpern) and key terms (constructor theory, resource theory) from the article into Google Scholar, or visiting the linked researchers' websites, will probably get you what you're looking for.
drallison 672 days ago [-]
David Deutsch as well. Chiara Marleto has a nice recent book, The Science of Can and Can't, which discusses briefly the Second Law in the context of constructor theory.
photochemsyn 672 days ago [-]
A little digging turns up this, which discusses their experimental work:
"(2022) Emergence of Constructor-Based Irreversibility in Quantum Systems:
Theory and Experiment, Marletto & Vedral"
That's related to this quote from the article:
> "But crucially, the theory also predicts that the row becomes even less able to do the reverse transformation from B2 to B1. The researchers have confirmed this prediction experimentally using photons for B and a fiber optic circuit to simulate a row of three qubits."
The experiment is rather complex, and would take a long time to analyze and understand in depth. They appear to be claiming that microscopic reversibility, a cornerstone of atomic-molecular kinetic theory, is being violated in their setup? Seems like a rather extraordinary claim but it passed peer review in a major journal so, worth digging into probably.
For those who want an introductory overview of the issue, this is a good one:
The laws of thermodynamics do not particularly concern themselves with the specific how and why of heat transfer, which makes sense for laws that were formulated before the atomic theory was fully adopted. They deal with the sum total of energy and heat transitions within a system and do not take into account the specific nature of heat transference on the atomic or molecular level."
In practice the second law is not all that useful although it's philosophically important I suppose. For example, a redwood tree growing from a redwood seed plus simple inputs (CO2, H2O, photons, NH3 etc.) is clearly a local decrease in entropy, as opposed to the dispersal of those elements when a redwood tree burns or falls and is digested by fungi, but since solar fusion is creating a much greater increase in entropy, there are no second law violations. Similarly the creation of heavy elements in a supernova would seem to be a decrease in entropy but the release of all that energy to the surrounding universe ensures no second law violations take place. In practice it's very difficult to do all the accounting for even moderately complex systems.
edgyquant 672 days ago [-]
Thank you for taking the time to dive into this
672 days ago [-]
atx42 672 days ago [-]
I've wondered why we don't just assume the arrow of time was reversed in the past? We know that in the past the universe was more organized (had less entropy), so it doesn't seem like much of a stretch from believing entropy started out that way. Wouldn't that make it easier things like life to form, etc. and simpler than reaching for alien life / multiverse theories? Or does it make it just as difficult? Just a lazy armchair physicist here, but maybe evening TO morning was the first day?
leephillips 672 days ago [-]
I can’t grasp the meaning of all of your comment, but I think it’s worth pointing out that there’s nothing in the Second Law that prevents life from evolving. We certainly don’t need to invoke multiverses for that. For systems out of thermal equilibrium, such as a planet embedded in a flux of energy from its star, entropy can and does decrease all the time.
An air conditioner (heat pump), for example, causes heat to flow opposite to the temperature gradient, pushing heat from your house into the outside, and increasing the temperature difference, thereby decreasing the entropy in your area. But you have to expend energy to do so, hence your electricity bill.
atx42 672 days ago [-]
Wouldn't the Big Bang represent a much higher level of entropy than now?
snowwrestler 672 days ago [-]
If the Big Bang actually happened, just prior to the bang would have been lowest state of entropy ever, as the entire universe existed in one single configuration.
zmgsabst 672 days ago [-]
Can you elaborate?
The universe has a collective state now, so I’m not seeing how scaling changes that.
dekhn 672 days ago [-]
They mean there was only one possible configuration of matter, like a die that could only ever roll to "1".
I like to think of entropy as "the universe prefers to have the greatest degrees of freedom available to it and there is a cost to reducing degrees of freedom". For example, to make a perfect crystal you have to cool to absolute zero (and even then there can be residual entropy!). Effectively, entropy = the logarithm of the number of ways a thing can be ordered.
If a thing can only be ordered in one way, then its entropy is minimized.
zmgsabst 672 days ago [-]
Well, there isn’t matter at that time — and your analogy to dice is exactly my point.
That instant could roll to anything from its current position (and according to many worlds, did) — the phase space of that object is huge, far beyond anything we normally see. That single object has all the complexity of every possible universe stacked into an infinitesimal region!
We’re in a colder universe than it was at that compacted moment, which I’m having trouble aligning with your explanation about crystals: we’re seeing the cooling towards absolute zero as a superheated droplet spreads out and freezes.
Defects forming in the crystallization of iron doesn’t imply the cold, defect laden iron has more entropy than the liquid drop it cooled from.
snowwrestler 672 days ago [-]
The classical way to think about entropy is a relationship between scales. There are a lot of ways for matter and energy to arrange as a “cloud of smoke.” There far fewer ways for matter and energy to arrange as you. You have lower entropy than a cloud of smoke.
Just before the Big Bang (again, if it occurred), there would have been a perfect 1:1 relationship. There was only one possible way to be, and everything was that way. Zero entropy.
It is basically a definition, not empirical. Entropy only increases; the universe has a definite and specific history; therefore the beginning of things must have been only possible exactly one way, the lowest-entropy way.
If this sounds weird and non-scientific, it is! Scientists don’t know if the Big Bang occurred. We have basically no evidence to support any theory of what things were like before inflation.
zmgsabst 672 days ago [-]
> There was only one possible way to be, and everything was that way. Zero entropy.
The issue I’m having is this:
There are a lot of states that initial singularity could have been in, leading to a whole diversity of universes.
How is it different than your smoke example?
snowwrestler 672 days ago [-]
There are a lot of possible singularities that could have existed, but only one did. (Again: if it did.)
At least, in the context of this universe. But that is the entire and only context available to us.
Whereas all sorts of different clouds of smoke exist, some probably even right this second just here on Earth.
zmgsabst 671 days ago [-]
> At least, in the context of this universe. But that is the entire and only context available to us.
But we know things take “multiple paths” — why does that not apply to the original singularity diverging into multiple universes?
> Whereas all sorts of different clouds of smoke exist, some probably even right this second just here on Earth.
But only a particular cloud exists — the one that comes out of your fire. You go from one log as it exists to one cloud as it exists… one state to one state.
We normally explain that as “there’s more potential ways to be a cloud than potential ways to be a log”, but that brings us back to my question…
Why do the potential other smoke clouds count as entropy in the case of smoke clouds but not potential other universes in the case of universes?
mensetmanusman 672 days ago [-]
It might be counter intuitive, but no. When you confine systems, the number of states that they can experience goes down quickly, so the Big Bang was an extremely low entropy point.
Ever since then the entropy of the universe has been increasing overall. We are about half way to the heat death of the universe from an entropy calculation point of view.
boole1854 672 days ago [-]
Actually, the early universe appears to have 'come out of the Big Bang' in an extremely low entropy state.
Does reversing time makes it "easier" to get life, than the current direction?
leephillips 672 days ago [-]
The direction of time is just a convention, and the laws of physics are (still, as far as I know) time-reversible, so this is really irrelevant. You need an energy source to make anything, including life.
pfortuny 672 days ago [-]
All except for (precisely) the heat equation, which is not time-reversible (there is a “loss of information” as time passes, very badly and roughly speaking).
No one can be sure, but I can effectively time reverse my AC in the winter for efficient heating. :)
Metaphysics isn't necessary (or helpful) in figuring out how we got here.
atx42 672 days ago [-]
Interesting. I guess it gets back to "first cause" argument. If life is organizational, does it require an organized process to form.
Isn't metaphysics the only thing we have to figure out how we got here? Provable or not? Even science agrees that there are true statements that are not provable.
leephillips 672 days ago [-]
We don’t need an organized process, intelligence, or intentionality to create order. We just need a flux of energy. We observe order arising without the intervention of an organizational force all the time, all around us. If you, like me, enjoy observing clouds, you may have noticed that they often form regular geometric patterns, often an area of stripes with a definite wavelength. This is a form of convection pattern that you can also observe in your kitchen. Nobody arranged the clouds that way; the order appeared “naturally”.
willis936 672 days ago [-]
Math says there are statements that are not provable. Scientific intellect is an entirely empirical set of descriptions of observations. All of its models have finite domains. Speculating outside of determined physical limits is metaphysics: pure fiction. It adds nothing of value and certainly has no bearing on important questions like the origin of life or the universe.
lrkr097 672 days ago [-]
Flying around the Earth super fast won’t literally rewind its state. It will erode current state but not restore old state. An accurate physical theory needs to account for events we can’t recreate influencing present state. That’s hard when objects and forces of influence can be separated by millions of light years at this point.
What these physicists are working on though is a math framework that might help simulate old state by being able to account for eons of time and vast distances traversed. So our intuition is as you say it’s just hard to accurately model what things would look like if such a reversing was to happen.
jonsen 672 days ago [-]
I think you might enjoy this lecture on the meaning of time:
https://journals.aps.org/prl/pdf/10.1103/PhysRevLett.128.080...
Figs 1 and 2 show the setup. Here's the conclusion:
> "This gives a clear proof of the compatibility of constructor-based irreversibility with unitary quantum theory, providing a frame for the emergence of thermodynamical irreversibility in quantum mechanics."
It seems like an extraordinary claim, but is very interesting. The theoretical framework (constructors) is a bit confusing. Hopefully this will be replicated.
I recall someone talking about something rather similar once, related to so-called chaos theory (now more generally called 'sensitive dependence on initial conditions'), which describes how deterministic systems can give rise mostly unpredictable phenomena (strange attractors etc.), in which indeterminate quantum states at the atomic level are the ultimate cause of the phenomenon.
Chaos is unpredictable but for a different reason than quantum systems: Chaotic systems are unpredictable because we can never measure the real world perfectly (even ignoring quantum weirdness) and that lack of precision in measurement coupled with nonlinearity means our measurement errors amplify quickly. Nevertheless, chaos is theoretically deterministic. It's just not practically deterministic (except in the coarse-grained statistical sense of saying "the system's position in phase space is somewhere on a strange attractor path bounded by X.")
Quantum systems are not even deterministic in theory, let alone practice. And ironically quantum systems are fundamentally linear. So the spookier theory in a sense is based on more tractable math.
It's important to note that the parts of quantum mechanics that are linear are also fully deterministic - the Schrodinger equation predicts the evolution of the wave function of a system with exact precision (as do the more complex equations of QFT).
The part that introduces non-determinism is the non-linear adjustment of the Schrodinger equation that you have to perform after a measurement - the Born rule, also called wave-function collapse. Here the complex-valued wave function of a system randomly takes a single real value at the moment of measurement. It does go back to its deterministic linear evolution afterwards though, but you have to perform this non-linear adjustment.
This non-linear adjustment can be interpreted in many ways, but without it QM doesn't agree with even the most basic experiments, so one way or another it is a vital part of the theory, rendering it non-linear overall.
When qbits have to be read, we convert them from complex to real and that step is both nonlinear and nondeterministic.
Chaotic orbits of whole moons are non-deterministic on quite short time horizons.
The chaos can act as a magnifier of quantum fluctuations so that they can produce sizable effects in the world around us. But we know that that can happen often. - Murray Gell-Mann
I guess the idea is that if you have a collection of particles in a system that displays sensitive initial dependence, then any quantum jiggling gets rapidly amplified.
[edit] As you note this doesn't explain the chaos phenomena, because a stable dynamic system would squelch any quantum weirdness rather than amplify it.
it's the same phenomenon underlying the wisdom of crowds (and the pricing function in fair markets): each individual can be wildly incorrect in an observation, but averaging many (independent) observations together reveals stable insight.
Some nonlinear DEs exhibit chaos. That’s a purely mathematical property. Whether any particular DE is a useful model of a particular physical system is a matter for the imagination, and either backed up or refuted by experiment.
Sure, but differential equations describe physical systems, and there is a canonical way to derive a quantum differential equation from a classical one by quantifying the classical Lagrangian using the Path Integral formulation. Giving a fairly natural distinction between the types of equations
> Whether any particular DE is a useful model of a particular physical system is a matter for the imagination, and either backed up or refuted by experiment.
This doesn't make sense to me. The Navier-Stokes equations are known to describe the classical behavior of water and are experimentally confirmed to predict things like trajectory. Their effectiveness has nothing to do with my imagination. If I write x=x' for the position vector of atoms in a fluid that will completely fail to describe anything physical.
Note that there is no Largrangian for the NS equations, by the way.
> Note that there is no Lagrangian for the NS equations, by the way.
I don't know much about fluid dynamics, but I was under the impression that Bennett derives the Lagrangian form in the book Lagrangian Fluid Dynamics
So this is primarily an interesting mathematical restatement at the quantum level. I don't know how useful this is, not having much knowledge about existing quantum information theory/thermodynamics.
isn't that the wolfram thing? Simple patterns create complex unpredictable patterns? And our universe are just made up a few very simple rules and in combination it created this complex universe. Yeah most ideas has being around for a long time, it is just now we have the technology to verify some of these claims.
I was also extremely puzzled by the dismissal of probabilistic or statistical models as being fuzzy or something. Probability has always been the domain of possibility -- this is basically the Jaynesian-style account of probability. It's always seemed to me to be the case that at some level statistics is about selective loss of information -- that is, as you go from descriptions at the particle level to group-level statistical accounts, you summarize (in a coding sense) the ensemble of particles' behavior to shed information required to maintain complete information about the ensemble. To make an analogy, if you wanted to know people's heights, at some point storing information about 1 million person's heights would become burdensome, so you'd replace it with a single number (e.g., the mean) instead.
That's not to say what's being discussed in the article isn't interesting, but it just seems like a different level of analysis or something? There's also different ways of representing amount of information in the quantum realm vs classical realm, so there's that, but it's also the case that a chamber of gas will generally involve a lot of molecules in many classical cases.
I also had the sense that there was probably something more fundamental that you could invoke that would integrate a lot of this, at the level of logic or computational theory or something. But maybe at this level the difference is sort of moot.
Perhaps you're just saying that probability is useful in some practical ways? Deutsch agrees (see video below). But if that's not all you're saying:
If the "domain of possibility" (which I interpret to refer to quantum mechanics and thermodynamics) can be described without the whole concept of probability, why not eliminate it? Why eliminate it, though? As Deutsch says "We don't need a reason: Physics likes to do without things"? [1]
You may say constructor theory adds new fundamental concepts. I don't know constructor theory so I don't know whether that's true. In cany case, what's not to like about probability concepts? In what way are they "fuzzy or something", as you put it? Deutsch points out here [2] (few minutes later in the same talk as the other link) a sense in which probability statements don't describe physical facts: for example "drawing to an inside straight is very risky" is logically consistent with any sequence of cards in a poker game. I think that's the "fuzziness" they want to eliminate from Physics. He then goes on to deal with some objections to that idea such as "in the long run, probability statements describe physical facts".
If the game of Physics is describing physical reality, and we don't need that specific kind of fuzziness, success would seem to me an advance in Physics.
[1] https://invidious.namazso.eu/watch?v=wfzSE4Hoxbc&t=13m12s [2] https://invidious.namazso.eu/watch?v=wfzSE4Hoxbc&t=14m18s
Could someone who knows the topic tell us (and by “us” I mean layman like me) if there is something to keep from this article? It brings so much context that after reading it I still have no idea what this “rewrite of fundamental law” is about.
There has always been something unsatisfactory about the second law of thermodynamics. Unlike many other laws in physics, which precisely state both what kinds of structures exist in reality and how they behave, the second law has a built-in lack of precision. Specifically, there are multiple statements of what the second law says, but they all suffer from referencing some concept which is inherently fuzzy at some scale. For example, they may refer to "heat" as part of the law, but the physical definition of "heat" is fuzzy at the microscale -- it only has a clear meaning for "large systems".
However, using a new physio-mathematical formalism called "constructor theory" the researchers have now found a way to state the second law of thermodynamics which is precise at all scales.
[1] https://journals.aps.org/prl/pdf/10.1103/PhysRevLett.128.080...
In the scientific sense, it is not enough to show that your system agrees with experiments to say it is true, you also have to show that the alternatives do not agree with experiments. Constructor theory is metaphysical because you'd get the same analysis whether you used it, ad-hoc intuitive thinking (in practice, people do not need instruction on universal principles like cause and effect), or some other alternative. All of the non-intuitive weight in that paper is borne by the QM stuff that exists independently of constructor theory, and in contrast, if you changed that the answers would start coming out wrong.
I am not proposing that this captures the essence of positivism, but one test that you can apply to check new ideas without going through the trouble of understanding them is to see whether they introduce any new conclusions that could not have been obtained under previous frameworks. If not, they may be metaphysical re-arrangements of existing ideas.
One of the points of the paper is that constructor theory makes such statements in a way that is compatible with other known laws. This is demonstrated via the experiment. So I'm afraid I do not understand your remark that "you could take the constructor theory language out and have the same paper". Taking the constructor theory language out would undermine a major point the authors are trying to make in the paper.
Yeah, but known laws are already compatible with themselves. You might make them more comprehensible by establishing a conceptual meta-framework that offers a consistent way to phrase statements about them, but that is clearly a philosophical task.
But I'll admit I don't understand constructor theory, and maybe it can hide any interesting discrepancies by overfitting.
"A scientific explanation is a statement of what is there in reality, and how it behaves and how that accounts for the explicanda." (emphasis added) [1]
Deutsch argues strongly against the view of physics which says that physical explanations are just about predictions of behaviors and not about what is really out there. See also his 1997 book The Fabric of Reality.
[1] https://www.sciencedirect.com/science/article/pii/S135521981...
Sure, you can say stuff about "what is there in reality", but I don't see any way one can verify such statements.
Everything humans have access to is mediated through their perceptions. Everything that we sense "outside" of us is to us a perception.
This includes the results of all the experiments ever done, all our interaction with tools and scientific instruments, and all our observations. They're all perceptions.
Our thoughts and feelings are also perceptions. The results of our deductions and inductions are perceptions.
Humans seem to be hermetically sealed from the "outside world" (if there is one) by our perceptions, and I don't see any way out of that... not through science, not through philosophy, not through religion, not through anything.
"Mach, Einstein, and the Search for Reality" https://dash.harvard.edu/bitstream/handle/1/37902464/2002383...
In short, late in the 19th century and early 20th century, the physics philosophy of Ernst Mach held great sway. Mach's viewpoint was the very viewpoint that you describe: the only things we know about are our own perceptions and proper formulations of physical theories consists at the fundamental level of descriptions of these things.
Mach had a significant impact on many physicists' understanding of what they were doing and also had an impact on the development of logical positivism in the field of philosophy.
Einstein accepted Mach's philosophy initially. But gradually he came to see it as untenable given his own work in physics. The article linked to above describes how and why his viewpoint changed over time.
However, if I gave you the impression that I was a Machian, this is a mistake.
According to this paper, Mach equated reality with sensory experience (ie. the experiences afforded us through the five senses: seeing, hearing, touch, smell, and taste.. and perhaps some others like proprioception), and that speculation about a world "outside" and independent of such sensations was unwarranted.
This is not my position, as I include in "perception", the perception of thoughts, theories, conclusions, intuitions, feeling and everything else considered to be "internal" -- which includes everything Einstein or any other physicist could ever base their own theories on.
Also, I don't object to speculation as to an "external", "objective" world, as Mach does. Science can come up with all sorts of interesting and possibly useful things through such speculation. I'm happy to read of such conjectures, and if some people want to engage in it more power to them.
However, do I not see any way we could ever verify whether such speculation right or wrong (or, to put it another way, whether such speculation corresponded to the way the world really was), because neither mathematical simplicity or the elegant fitting in to other, larger theories (as Einstein seems to have favored) means that the speculation has anything to do with the "external" world as it is (or even that there is an external world).
Since we seem to be limited to our own experience, the belief that an "external" world exists and that we have any (direct or indirect) way to access it seems to be nothing but faith.
The scientific method is about predictions and testability of such predictions, therefore it implies behavior. Nothing more is needed for a scientific theory. To test for existence within the scientific method would require to test for non existence too, a clear paradox, which is the reason why ontological statements ("X exists") are meta-physical in nature.
How far do you take that stance?
If a biologist says a particular bacterium exists, are they doing philosophy, not science?
Edited to add another interesting case: creation "science" organizations like Answers in Genesis say that the theory of evolution is a philosophy not science since it describes non-observables in the past (events in the past are inherently not observable now). When a paleontologist says a particular animal once existed in the past, would you say they are they doing philosophy not science?
When a biologist says that a bacterium exists, they are making a descriptive statement, not an ontological one. To say that a bacterium exists is just a convenient proxy to refer to a specific aggregate as a whole. A bacterium is always reducible to its constituents, therefore has no ontological status nor existence in and of itself: in other words, if the molecules which the bacterium is made of would relinquish their structural integrity, the bacterium would cease to "exist" too.
Therefore I would agree with you that it's totally legitimate for a scientist to claim that a bacterium exists, in so far as we understand this claim as a descriptive proxy. A shorthand, essentially, to cluster together specific configurations of the underlying, supposedly more real, "stuff". And this is needed, if we want to make any progress. We have to communicate, effectively and succinctly. Science is precisely about _descriptions_ and predictions of behavior. Same goes with the paleontologist example.
It's a completely different matter instead to claim ontological status, that is, to say that something exists in and of itself, irreducible to something else, as a building block of the universe. This is a philosophical stance per-force, since the scientific method cannot follow us there. It's meta-physics by definition, and there are different rules at play there. Not Popper's falsifiability, but empirical adequacy, internal consistency, explanatory power, etc.
So when scientists claim that the ontological primitive is, e.g., the quantum field, or "energy" or anything really, they are stepping soundly and firmly into philosophy. Which is totally fine to do. We just need to be aware of this, and start to apply different rules to evaluate the validity of the claims.
Edits are typos.
However, note that in my original post I didn't claim that physical laws describe what exists at an ontologically fundamental level, just what exists in reality. And by that I mean "exists" in the non-subjective way that, say, hydrogen atoms exist. I wouldn't mean to imply that hydrogen atoms are irreducible.
Does Deutsch and his research group make claims to know what is ontologically irreducible? Not as far as I know, but I am open to correction!
I actually envisioned a similar thought experiment that could be done with a very complex quantum computer (say millions of qubits) to test whether the universe is actually keeping tabs of all these states or if it gives up at some point and produces irreversibility. But I guess if you'd formulate it like that the academic community wouldn't take you very seriously anymore. I'd wager Deutsch's actual motivation might be along this line, as he seemed very interested to find a way to test the many worlds interpretation against other interpretations of quantum mechanics.
See https://www.sciencedirect.com/science/article/pii/S135521981...
http://blog.rongarret.info/2019/07/the-trouble-with-many-wor...
The multiverse interpretation accounts for cases where the probabilities of each distinct observable outcome are equal according to the Born rule. But it cannot account for cases where the Born rule says there are non-equal probabilities for the possible outcomes.
This is a fairly common problem which critics of the multiverse interpretation raise. See [1] for another example where a critic raises that problem in more detail (with examples).
Deutsch and others have responded. In short, the reply is that the mathematics of the Schrodinger equation describes a density of universes in the continuum which is the multiverse, not a finite set of distinct, interacting single universes. This undercuts the critics' argument. See [2] for an example of this kind of response.
[1] https://arxiv.org/pdf/0810.2657.pdf [2] https://www.pnas.org/doi/pdf/10.1073/pnas.1324238111
> This undercuts the critics' argument.
No, it doesn't, for all of the reasons given in the blog post and the subsequent discussion. The Born rule is an observable fact of nature, and you cannot derive it from the Schroedinger equation, Duetsch and Wallace's claims to the contrary notwithstanding.
Edited to add my understanding: if the cardinality of the multiverse is uncountable, one has to talk about densities of branches. In that scheme, the problems which branch counting introduce do not arise.
It doesn't (see below) but even if it did...
> doesn't that undercut the argument in the blog post?
No. It's really very simple: the SE is linear. The BR is not. You can't derive a non-linear result from a linear equation.
The only thing that is even a tiny bit complicated about this is figuring out where the Deutsch-Wallace argument fails, because Deutsch and Wallace are not idiots and they hid their questionable assumption very well.
So... why does the SE not describe a "density of universes"? It's because the SE does not describe universes at all! Deutsch and Wallace specifically acknowledge this, which makes their argument all that much more untenable. The SE describes the behavior of a function called "the wave function". The wave function is not a description of anything physical. The domain of the wave function is not physical space, it is configuration space. This is the reason entanglement seems weird to us, because we live in physical space and the wave function does not.
In order to get anything physical out of the wave function -- particles, universes, anything resembling our classical reality -- you have to add some additional assumption, and any additional assumption that you add will be in some sense isomorphic to the Born rule. There is no way around this. Trying to derive the BR from the SE is like trying to prove Euclid's fifth postulate.
This is circular: you're assuming that realism about the wave function is incorrect to argue against the (tragically misnamed) many-worlds interpretation, but the only substantive claim the MWI makes is that realism about the wave function is true.
No, I'm not assuming it. I simply observe that the SR by itself does not describe the outcomes of experiments. And I'm not even talking here about quantum experiments. The SR cannot by itself even account for the fact that we live in a three-dimensional space or that there is an arrow of time. And it certainly cannot account for the Born rule, which is also an experimentally observed fact.
> the only substantive claim the MWI makes is that realism about the wave function is true.
Yes, I know. The problem with this is that it is violently at odds with observations that every human makes every moment of their lives, and it does not account for the Born rule.
Other than that there's no problem with it.
No formalism describes the outcome of experiments by itself. There always needs to be an interpretation attached, and our choice of interpretation is always underdetermined by observation.
> The problem with this is that it is violently at odds with observations that every human makes every moment of their lives
Like all other interpretations of all other physical theories, MWI isn't and can't be at odds with our observations alone, because our observations alone aren't enough to imply any particular relationship between observations in general. And like all other interpretations, it is and must be at odds with competing interpretations, and in particular with the folk-physics model we're naturally inclined to use. But this means very little, because folk-physics is wrong, and this far from the paleolithic savanna, it's wrong in ways that matter. Intuition is evidence, but it's not particularly strong evidence, and it's certainly not proof.
> and it does not account for the Born rule.
Sure, it still needs a separate account of subjective probability. But so does everything else: it's just a little harder to notice in the classical regime, since the relevant probabilities are always either 1 or 0.
That is so wrong as to border on the ridiculous. The quantum mechanics formalism, i.e. the SE plus the BR together describe the outcomes of experiments. No interpretation is required. You just turn the crank on the math and predictions pop out. There is a reason that "shut up and calculate" is a thing.
The only reason people talk about interpretations at all is that the structure of the math is violently at odds with our intuitions and subjective experience, and this leaves some people feeling queasy and wanting an explanation. There's nothing wrong with that. But it has absolutely nothing to do with the power of the formalism to make predictions, which is completely independent of any interpretation. (And this is true of classical mechanics as well.)
Further, the author of that paper, Frank Tipler, is a notorious nutjob: https://en.wikipedia.org/wiki/Frank_J._Tipler.
I don't have time at the moment to review the paper in detail to locate any potential flaws, but I wanted to flag this for others. I think there is substantial reason to be skeptical.
I also agree there is no consensus that the problem of getting Born's rule from the Schrodinger equation has been solved to everyone's satisfaction.
Andrew Gelman calls it a "tabloid": https://statmodeling.stat.columbia.edu/2014/06/17/hurricanes....
I will check the paper out though.
[0] https://www.science.org/content/blog-post/no-more-prearrange...
Edit: I read the paper. It is borderline crank work. There are decent arguments for the Born rule in MWI but this is not one of them. (I like David Wallace's book, for example.)
One basic problem is that using symmetry arguments to justify the assignment of subjective probabilities (cf. page 4, right column) is going to fail when there isn't a symmetry available, for example for irrational probabilities.
A deeper issue is the implicit assumption that one can coarse-grain into discrete "effectively distinguishable universes" for the purpose of doing these counting arguments. There are significant conceptual difficulties doing this, and this claim is much stronger than the bare MWI. Wallace discusses this at length; see for instance the citations in the blog post you responded to.
This paper introduced constructor theory: Deutsch, D. Constructor theory. Synthese 190, 4331–4359 (2013). https://doi.org/10.1007/s11229-013-0279-z (Download preprint version: https://arxiv.org/abs/1210.7439)
Then this paper applied constructor theory to form a constructor theory of information: Deutsch David and Marletto Chiara 2015 Constructor theory of information Proc. R. Soc. A.4712014054020140540 (Download: https://royalsocietypublishing.org/doi/10.1098/rspa.2014.054...)
Then this paper applied the constructor theory of information to probability: Marletto Chiara 2016 Constructor theory of probabilityProc. R. Soc. A.4722015088320150883 http://doi.org/10.1098/rspa.2015.0883 (Download: https://royalsocietypublishing.org/doi/10.1098/rspa.2015.088...)
Finally, this paper applied the constructor theories of information and probability to thermodynamics: https://arxiv.org/pdf/1608.02625.pdf (not yet published in a journal, to my knowledge)
For what it's worth, the Deutsch "Constructor Theory" stuff has its own homepage at https://www.constructortheory.org/
There is also a Wikipedia page that has a list of references and external links.
https://en.wikipedia.org/wiki/Constructor_theory
https://journals.aps.org/prl/pdf/10.1103/PhysRevLett.128.080...
"(2022) Emergence of Constructor-Based Irreversibility in Quantum Systems: Theory and Experiment, Marletto & Vedral"
That's related to this quote from the article:
> "But crucially, the theory also predicts that the row becomes even less able to do the reverse transformation from B2 to B1. The researchers have confirmed this prediction experimentally using photons for B and a fiber optic circuit to simulate a row of three qubits."
The experiment is rather complex, and would take a long time to analyze and understand in depth. They appear to be claiming that microscopic reversibility, a cornerstone of atomic-molecular kinetic theory, is being violated in their setup? Seems like a rather extraordinary claim but it passed peer review in a major journal so, worth digging into probably.
For those who want an introductory overview of the issue, this is a good one:
https://www.thoughtco.com/laws-of-thermodynamics-p3-2699420
> "Kinetic Theory & the Laws of Thermodynamics
The laws of thermodynamics do not particularly concern themselves with the specific how and why of heat transfer, which makes sense for laws that were formulated before the atomic theory was fully adopted. They deal with the sum total of energy and heat transitions within a system and do not take into account the specific nature of heat transference on the atomic or molecular level."
In practice the second law is not all that useful although it's philosophically important I suppose. For example, a redwood tree growing from a redwood seed plus simple inputs (CO2, H2O, photons, NH3 etc.) is clearly a local decrease in entropy, as opposed to the dispersal of those elements when a redwood tree burns or falls and is digested by fungi, but since solar fusion is creating a much greater increase in entropy, there are no second law violations. Similarly the creation of heavy elements in a supernova would seem to be a decrease in entropy but the release of all that energy to the surrounding universe ensures no second law violations take place. In practice it's very difficult to do all the accounting for even moderately complex systems.
An air conditioner (heat pump), for example, causes heat to flow opposite to the temperature gradient, pushing heat from your house into the outside, and increasing the temperature difference, thereby decreasing the entropy in your area. But you have to expend energy to do so, hence your electricity bill.
The universe has a collective state now, so I’m not seeing how scaling changes that.
I like to think of entropy as "the universe prefers to have the greatest degrees of freedom available to it and there is a cost to reducing degrees of freedom". For example, to make a perfect crystal you have to cool to absolute zero (and even then there can be residual entropy!). Effectively, entropy = the logarithm of the number of ways a thing can be ordered.
If a thing can only be ordered in one way, then its entropy is minimized.
That instant could roll to anything from its current position (and according to many worlds, did) — the phase space of that object is huge, far beyond anything we normally see. That single object has all the complexity of every possible universe stacked into an infinitesimal region!
We’re in a colder universe than it was at that compacted moment, which I’m having trouble aligning with your explanation about crystals: we’re seeing the cooling towards absolute zero as a superheated droplet spreads out and freezes.
Defects forming in the crystallization of iron doesn’t imply the cold, defect laden iron has more entropy than the liquid drop it cooled from.
Just before the Big Bang (again, if it occurred), there would have been a perfect 1:1 relationship. There was only one possible way to be, and everything was that way. Zero entropy.
It is basically a definition, not empirical. Entropy only increases; the universe has a definite and specific history; therefore the beginning of things must have been only possible exactly one way, the lowest-entropy way.
If this sounds weird and non-scientific, it is! Scientists don’t know if the Big Bang occurred. We have basically no evidence to support any theory of what things were like before inflation.
The issue I’m having is this:
There are a lot of states that initial singularity could have been in, leading to a whole diversity of universes.
How is it different than your smoke example?
At least, in the context of this universe. But that is the entire and only context available to us.
Whereas all sorts of different clouds of smoke exist, some probably even right this second just here on Earth.
But we know things take “multiple paths” — why does that not apply to the original singularity diverging into multiple universes?
> Whereas all sorts of different clouds of smoke exist, some probably even right this second just here on Earth.
But only a particular cloud exists — the one that comes out of your fire. You go from one log as it exists to one cloud as it exists… one state to one state.
We normally explain that as “there’s more potential ways to be a cloud than potential ways to be a log”, but that brings us back to my question…
Why do the potential other smoke clouds count as entropy in the case of smoke clouds but not potential other universes in the case of universes?
Ever since then the entropy of the universe has been increasing overall. We are about half way to the heat death of the universe from an entropy calculation point of view.
See https://www.ws5.com/Penrose/
See https://math.stackexchange.com/questions/628720/why-heat-equ...
Metaphysics isn't necessary (or helpful) in figuring out how we got here.
Isn't metaphysics the only thing we have to figure out how we got here? Provable or not? Even science agrees that there are true statements that are not provable.
What these physicists are working on though is a math framework that might help simulate old state by being able to account for eons of time and vast distances traversed. So our intuition is as you say it’s just hard to accurately model what things would look like if such a reversing was to happen.
https://m.youtube.com/watch?v=-6rWqJhDv7M