Hacker News new | past | comments | ask | show | jobs | submit login

Okay, so, let me try to enumerate these.

(1) The hardest part of quantum mechanics to understand is intimately tied to our nature as subjects. Quantum mechanics is an entirely deterministic theory about waves in superpositions which even explains how, why, and when it disappears and averages out into classical probabilities. Except it doesn't tell you why we don't see your television as 50% in your bedroom and 50% in your living room. For that we start to invent things like many universes, half of which have a TV in your bedroom and half of which have a TV in your living room -- or Penrose's idea that perhaps some magical interaction not yet observed forces something as large as a TV to "make up its mind" due to its gravitational warpings. It's just the fact that your subjective experience doesn't see this TV as half in one place, and half in another, where the wavefunction description fails. The rest is clean, deterministic, and precise.

(2) The second hardest part of consciousness to understand is something we basically take for granted in the hardest part of quantum mechanics. Compatibilist theories of free will make you feel warm and fuzzy but they don't really feel like they address the fundamental mystery: How can a deterministic system be said to "choose" something when that's a fuzzy airy-fairy notion and the actual outcome was written down in the book of history 13.7 billion years ago, full stop? Why do we feel like we've got a million options open when in fact this is false, false, utterly false, moronically-out-of-touch-with-reality crazy false? But in quantum mechanics, the idea that a quantum system has many options open, and can even "sniff around" over the whole space, and that it might make a difference if it does or not -- these things are very acceptable even if you take a deterministic approach to wavefunction collapse. You could maybe in principle go from "warm and fuzzy but I don't get it" to "weird and fuzzy but I think I get it," if the brain is intrinsically quantum.

(3) The hardest problem of consciousness -- how we reconcile our touchy-feely alive experience of red with a machine's cold unemotional ones and zeros -- is very easy to solve by saying "we aren't a machine; subjectivity is something more than a program." In fact, since programs can carry out programs, if consciousness is mechanical we might well ask, how do you know that your brain isn't a different conscious process which happens to be carrying out your own conscious process, and might someday soon get bored and kill you out of its own boredom and desire to live authentically? (This absurdity is also at the heart of Searle's Chinese Rooom argument.) The problem is that machines are such a good model that we have no idea what to replace them with. If quantum systems could really feel in a way that machines really can't, this would solve the problem neatly.

(4) One minor problem in consciousness is about how I'm able to hear things as I am seeing them -- multithreading would suggest that we have multiple consciousnesses doing the separate tasks separately; multitasking would suggest that we switch between these things so fast that it gives an illusion that they all come together, but I can't really do both at once. Quantum mechanics offers a tantalizing hint; for example, in a superconductor, two electrons of opposite momentum "get together" to form a larger particle called a Cooper pair. (You're advised not to think too hard about how they "stay together" with opposite momenta, because It's All Just An Excitation Anyways.) That larger particle can even do something which electrons never can -- it can form a Bose-Einstein condensate: it can conduct current without resistance.

(5) Everybody who watches Star Trek wonders what would happen if someone accidentally forgot to dematerialize you before re-materializing you down on the planet -- which one would be the right "you"? Quantum systems cannot be cloned in this way. The closest we can get is to identically prepare a system -- i.e. take another baby with the same genetic code and give it all of the same life experiences, and kill the ones who don't make the same choices in the same order. Without one of those, you can't clone an arbitrary unknown system into two copies without measuring it and thus forcing it into a classical state. If consciousness is quantum, then the classical state is dead. But if consciousness is just a program, then we're faced with a much scarier situation: in the transporter, you will actually die, but that's okay, because we'll create an impostor who will sleep with your wife for you -- in other words, it's okay because we'll never know the difference, and you'll be dead.




Just in case you are serious, or people believe you're serious:

1) Quantum mechanics operates on the quantum level. Macro objects, while composed of countless quantum systems, and thus are the result of the quantum systems of probabilistic particles and waves collapsing and such, are just that: the result of the quantum systems. This isn't subjective experience; it's an underlying behavior baring forth a different level of behavior. Just as it's possible, but doesn't make sense to describe Tetris in terms of electrons flowing through copper and semiconductors, one can describe the quantum behavior of what the TV is constructed of, but that wouldn't relate a bit to the meaningful object you watch shows on.

2) Probability on the quantum scale has nothing to do with the brain making choices; even if understanding quantum mechanics turns out to be essential to understanding consciousness, the brain can't change the laws of physics; it can't decide that certain waveforms will collapse in a way that appeals to it. By the same token, if you're a rational thinker, you'd understand that yes, if one develops an accurate enough model of the universe, the base stuff it's made of, you can predict the future. That's literally what science is; developing models which will predict how certain systems behave in a given condition. However going back to bullet 1, the concept of "free will" has nothing to do meaningfully with the base stuff of the universe. It's technically possible to be described in such terms, but that's not where decisions happen.

3) Again, going back to rationality; humans are part of the same deterministic system, operating under the same deterministic rules as everything else that exists. Sure subjectivity feels different than what we imagine a computer's (non)experience to be, but that's just a result of the conscious process (and the fact that most people don't empathize with machines).

4) I honestly don't even understand what you're trying to say here :\

5) This has nothing to do with anything at all but... you aren't the atoms that comprise you. You aren't even the -cells- that comprise you. "Seven years from now your body will be made of entirely different cells than it is now" and all that.


1) The problem that quantum mechanics has with consciousness is that you can connect macro objects (a cat) with the quantum effects on the quantum level (detector), so that we should actually observe quantum effects on macro objects. That's why Schrödinger's Cat is a paradox.

2) You can't really predict the future of a non-trivial closed system from within the closed system itself - no matter how you would "calculate" the future, you would never be able to include the state of your calculator/computer into the prediction (perfect compression does not exist). I.e., if I make a machine that predicts the future, I know the future, I can act to change it, so the future is different... and the machine does not predict the future, really.

In a way, the universe is the computer that is calculating its own future all the time.


> In a way, the universe is the computer that is calculating its own future all the time.

Reminds me of this book:

http://www.amazon.com/Programming-Universe-Quantum-Computer-...

> Is the universe actually a giant quantum computer? According to Seth Lloyd—Professor of Quantum-Mechanical Engineering at MIT and originator of the first technologically feasible design for a working quantum computer—the answer is yes. This wonderfully accessible book illuminates the professional and personal paths that led him to this remarkable conclusion.


I agree with most of what you said except "Seven years from now your body will be made of entirely different cells than it is now". Some neurons in the brain are never replaced http://askanaturalist.com/do-we-replace-our-cells-every-7-or... (couldn't find a better reference)


Maybe, but still their atoms are always on the move...


(1) There is no firm distinction between the "quantum level" and the rest of our universe; a "Heisenberg cut" needs to be placed, but where it occurs doesn't matter. And quantum mechanics is multiparticle, so rather than "countless quantum systems," these "macro objects" are a quantum system -- just one. There is a wavefunction for the universe, in principle, if you allow the clean deterministic unitary part of QM to do its thing.

But the quantum behavior does really matter. The clean part of QM says that if there's an accumulation of events which would transfer my TV into my bedroom with probability 50%, my TV is described by the state matrix ½ |L><L| + ½ |B><B|, where |B> is its state in my bedroom and |L> is its state in my living room.

But we don't see that state. We see |L> or we see |B>. That is the "hardest part" of quantum mechanics, where one insists that the above superposition describes many worlds or that there is a nonunitary wavefunction collapse or whatever.

(2) I don't require that the brain change the laws of physics or decide that certain waveforms collapse in any particular way. However, for reference, quantum mechanics does work that way, so get used to it. A system "over here" can indeed "decide" how a system "over there" will collapse.

If I have photons over here in a fiber-optic cable, I can perform a measurement which will change the result of a double-slit experiment over there. I can choose whether the people over there will see an interference pattern or not. In fact, I can choose whether it will happen after their detectors have already measured their photons. We've done this experiment, it has been confirmed, and if it weren't confirmed, QM would just be flat out wrong. [Theory details: start with (|00> + |11>) ⊗ (|0> + |1>), apply a CNOT from qubit 3 to qubit 2, then choose to measure qubit 1 in either the |0>, |1> basis or the |0> + |1>, |0> − |1> basis to destroy or restore the interference pattern at qubit 3. It's called a "delayed-choice quantum eraser."]

We gave up that very sentence, "if one develops an accurate enough model of the universe, the base stuff it's made of, you can predict the future", with the advent of quantum mechanics -- at least, in the way I think you intend it. Hey, here's a nice quantum system, it's in the state |0> + |1>. You measure it in the computational basis. No matter how accurate your model is, you cannot predict which one comes out. If you could, you could do experiments to prove that nature was self-contradictory.

(3) The problem is precisely that the "results of the conscious process" don't appear to be mechanical, such that we can't empathize with machines. It's not just "we don't," but that machines are so precisely understood, and so simple, that you would have to invent a "ghost in the machine" and believe in weird supernatural crap to have them really feel anything. Hell, you probably think, for all your talk of rationality, that I really believe in ghosts -- and it's for precisely this reason. Machines seem to require a "ghost" inside them to feel.

What I want to say is different. If this supposition is correct, then applying it back to our own case poses a trilemma. Either (a) we don't really feel anything, or (b) we are not machines, or (c) we're stuck with weird supernatural crap. Of these three, (b) is the "easy way out;" abandon functionalism. It comes with its own problems about what we replace it with; functionalism is not easily replaced. There is an alternative (d) which is to not take the problem seriously, which many influential philosophers including Dan Dennett have made a reasonable option, but it seems unsatisfyingly to fall into the attitudes of (a) and (c) when actually practised.

(4) Quantum systems sometimes form higher-order excitations which escape the parallel and the serial. Electrons in parallel only flow with resistance. Electrons in series also only flow with resistance. A Cooper pair is a two-electron excitation which flows without resistance, even though its constituent electrons could only flow with resistance. It requires a higher-order description. That is very rare in mechanics.

(5) Take it one step further: classically, you aren't even the pattern of atoms that comprise you. That's what the Star Trek example proves. We could create two copies of this pattern classically, and there would suddenly be two distinct people in the universe. If someone pointed a gun at you and said, "hey, we've got to kill you, we created this identical doppelganger of you down on the planet, and now we're convinced that he's the real you," that would seem awfully disturbing to anyone. The only reason they go along with it in Star Trek is because we kill them before they can react.

But quantum mechanics offers something fundamentally different. You could be the quantum pattern. Or, at least, the Star Trek thought experiment wouldn't suffice to disprove it. That's why one might be hopeful that there is something quantum about mind.


> If quantum systems could really feel in a way that machines really can't, this would solve the problem neatly.

Late to the party, but pretty sure that's the situation we're in.


I'm not sure whether you're being satirical or whether you're explaining viewpoints other people or whether you're representing your own opinion. In response,

(1) It's deterministic w.r.t. the wave functions, but you can't really measure wave functions, you can only measure classical things like momentum or energy. (A measurement is actually an interaction, of course.) If you manage to give your TV a wave function 50% in the bedroom and 50% in the living room, then you decide to watch TV, you've just given yourself a wave function that is 50% in the bedroom and 50% in the living room, and 100% correlated with the TV's wave function. A wave function can't be measured because if you try to measure it, all you're really doing is correlating yourself with the wave function. So the real problem is that it's hard to reconcile our description of quantum mechanics with our experience of it. We describe QM as if we're outside Schroedinger's box, but we experience it from the inside. "Many worlds" describes it from the inside, but people don't like the many worlds interpretation exactly because it doesn't privilege one of the worlds as "real".

(2 & 3) Theories of free will were kind of worn out during the 17th century. Associating free will with quantum mechanics is kind of like shoving the Christian god into the cracks in evolutionary theory. There's little actual evidence that the brain takes advantage of QM decoherence on any scale worth appreciating. QM consciousness theories seem to have forgotten all of the progress during the age of enlightenment (David Hume, for example) wrt the definition of free will itself, and the phrasing of questions about free will in such a way that the discussion can be productive. The discussion about "qualia" in philosophical circles is equally unproductive. They call it "qualia" because they don't want to be called dualists, which is understandable.

(4) I'm not sure how this has anything to do with QM. The fact that our brain can do to things at once (which it demonstrably can) doesn't have anything to do with quantum mechanics. A car can do things that engines or wheels never can, but you'd never call it quantum.

(5) QM provides an "out" for objects which are cloned in the classical sense. In QM, there is no such thing as identity; two electrons (for example) are interchangeable. This means that if you prepare a classical clone of yourself but the two clones get mixed up, from a QM perspective the question of identity is meaningless. So if you say X is the clone and Y is the original and Y is dead, the labels "clone" and "original" are actually arbitrary and you could relabel X as the original and Y as the clone, so the clone is dead and the original is alive.

It's like you have a bucket of electrons, and you paint one of them orange so you can remember it later. Can't do that in QM. Even if you put one electron in a magnetic trap, there's no saying it doesn't swap places with one of the electrons still in your bucket. You can't paint one copy of yourself as "clone" and one as "original". Of course, two copies of a person (neither privileged over the other as "original") don't stay interchangeable for long.

Summary: Many philosophers and physicists alike have a warm and fuzzy conception of consciousness, but they're just being noisy and some folks are actually making progress. Warm and fuzzy is old news, and quantum is just the latest flavor of warm and fuzzy.

Daniel Dennett is an example of someone worth listening to about cognition, though I'm not saying he's right.


(1) I believe we just said the same thing. I agree. The hard part to understand is this "we describe it that way, we live it this way."

I should hasten to add that there is a slight problem with what you've said, called the "preferred basis problem," but for reasons which I don't want to get into here, I suspect that the unitary part of QM is also able to handle this to some extent.

(2/3) Even if theories of free will were worn out in the 17th century, the feeling of free will is pretty hard to ignore; it's just this feeling that there are many options open, that you have a meaningful choice between many possible futures. It's the same thing with qualia. Sure, you can get a lot out of red by saying "color is a 3-tuple of integers, the red predicate holds when, after subtracting out a background tuple, the first integer is larger than the other two by a substantial degree." We render all sorts of red images this way, we can build computers which identify red things in photographs, and so forth. But that's nothing like how we experience redness, and it's extraordinarily hard to express what we experience in that vocabulary.

(4) I guess I'm getting at the fact that the rules for composing quantum systems are much more important than the nondeterminism. To the extent that a car has multiple functions, like the ability to play music and the ability to drive somewhere, we generally either split it up into components acting in parallel, or split it up into components which act in sequence. Similarly, our conscious experience has many perceptual functions, but our core consciousness itself does not seem to be divisible in that way.

It's like the old crap about the "inner theater": seeing something ends up "displaying it on your inner screen". The big problem with the inner theater is that I still have to put a unified you inside of it; it isn't sufficient to separate your senses into different theaters. The composition rules for neurons just don't seem to create the unified whole that we experience.

(5) That sounds to me like a cop-out. Sure, we've got two electrons and we can't tell which is which. But you'd need a much more radical application of quantum mechanics to say "therefore, there is only one electron, really." You would still seem to be confronted with the fact that there are two consciousnesses which you can't distinguish -- you would just have the added problem that you can't tell which was the "original" who you are "supposed" to destroy, and which was the "copy" who is permitted to live, because that's an arbitrary distinction. But it would seem pretty radical to say that, even for their first moments of life, there is one consciousness spanning two bodies such that killing one of those bodies does not amount to murder. Even though you can't tell these two apart, it would seem straightforward on most functional theories of mind to say that there are indeed two of them.

I agree with your summary in general -- I especially like one comment by Searle on free will: "Now, I dread saying this because, Searle's Third Law is, whenever philosophers talk about quantum mechanics, what comes out is hot air at best. And by the way, that applies to a lot of the things physicists say about quantum mechanics too."

But I was specifically answering the question, "why are some researchers so quick to blame our failure to understand consciousness on quantum mechanics when we don't yet understand the classical parts of the brain?" I wasn't being satirical. Some of these points come from other people; some of them come from me, but the aggregate point is that it's not just a "latest" flavor of warm and fuzzy: it's a flavor of warm and fuzzy uniquely suited to the problem and disproportionately confirmed by experiment to be a key part of the universe's workings. I agree that it's so hard to reason about and understand quantum mechanics that mechanical models are not just "actually making progress", but are much easier to grapple with and prove or disprove.

It's like if you ask me why us physicists ever accepted string theory, with its difficult mathematics and huge multiplicity of dimensions and arbitrary geometries over those dimensions. We've been interested in it because gravity just falls out. In quantum mechanics, some of the building blocks of warm-and-fuzzy consciousness also just fall out. The problems with both string theory and quantum consciousness are that the rest of the model is considerably more difficult, even though the hard problems of today just seem to have a promising solution in those terms.


(3) My first problem is that most definitions for qualia I come across are back doors for dualism. The second problem is that I'm not sure there's any point to discussing purely subjective phenomena, since a discussion by its nature requires objective description.

(4) I don't have an inner theater, or if I do, I don't call it that. I have no idea what you are talking about. I don't know how this relates to a supposed indivisibility of consciousness (I mean... how are we supposed to know it's indivisible, since we lack the technology to cut a brain up and split it between bodies, not to mention the ethical issues? Am I supposed to trust you that it's indivisible? And what about the evidence from split-brain patients?)

If anything, I perceive a very disjointed self. For example, I feel like I have a distinctly single-threaded audio processor which can run independently of other functions. I can play the piano and talk, but not at the same time, and when I'm doing other things quietly I get free background music in my head without using an iPod. (I wouldn't call it inner theater.)

One reason we can't understand brains the way we understand cars is because the human brain is more complex by many orders of magnitude. So saying our "consciousness itself does not seem to be divisible in that way" is not very convincing. Microprocessors also don't seem to be divisible that way, but I've been assured that they are. And another important difference is the processes that designed cars, microprocessors, and brains. Part of the design criteria for cars and microprocessors is that they be easy to understand from understanding their parts. The car is designed to be understood.

A better comparison might be circuits designed using powerful optimization algorithms. A number of researchers have designed circuits using general purpose algorithms (such as genetic algorithms or swarm algorithms). For all but the most basic end goals, the circuits are practically incomprehensible. One researcher designed an analogue circuit to compute the cube of an input voltage, the resulting circuit defies understanding even though its components are clearly simple transistors. Another researcher made a signal analysis circuit using an FPGA, the resulting mess was a technological marvel but not something you could understand by sitting down and looking at the schematic. Given that the human brain is probably at least ten orders of magnitude more complicated, and created using an equally powerful optimization algorithm, there's no need to bring in quantum mechanics to explain why we don't understand the brain.

I think the major problem here is that too many people I meet have a narrow view of what it means to "understand" something. We're used to understanding (a) simple things, like thrown rocks and (b) complicated things for which there exists significant neural machinery to analyze, such as used car salesmen. As the things we try to understand become more and more complicated, the way we understand them becomes more and more like the Chinese room.

(5) Never said there was one electron. I said that the two electrons were interchangeable. I'm saying if you actually could clone someone to the extent necessary for Star Trek teleportation but forgot to dematerialize the original, then the label of "clone" and "original" is arbitrary. To me, it seems a cop out to assume that everyone agrees that one is the clone and one is the original.

Could you clarify what you mean by "confirmed by experiment" and "key part of the universe's workings"? The problem here is that the subject of the sentence is "it", so I'd like to know exactly what is confirmed by which experiment, please.

I'd also like to hear how consciousness just "falls out" of quantum mechanics, since my current understanding of Newtonian mechanics is sufficient to me. My theory is that I am a machine whose output is composed of deterministic parts and thermal noise, and I have yet to encounter experiences not explained by this theory.

As for string theory, its acceptance is hardly universal.


(3) To solve your first problem, remember that a definition is independent of its rhetorical purpose: that is, it tells you what we're talking about, but it does not force or "backdoor" any sort of metaphysics upon you. To solve your second problem, realize that you're committing a fallacy of equivocation: you're assuming that because something is ontologically subjective -- that is, it only makes sense in the context of a subject which experiences it -- that it is epistemically subjective -- that is, it can only be known, talked about, and reasoned about, by that subject who experiences it.

(4) Just look at the words you chose: "I perceive a very disjointed self." You chose the singular 'I' and the singular 'a self'. As a subject, you would like to be just one subject -- you would like all of the parallel efforts of your different neurons to identify one "me". But then suddenly you're astonished that all of your experiences must be described as unifying into a single whole, which can't simply be treated as separate senses, but come to you as part of one whole conscious experience. That in turn astonishes me.

If you believe Sperry's ideas on his split-brain patients, then yes, there is evidence that to divide the brain creates two conscious processes, simply by cutting off communication between parts of the brain. I guess you could add that to the list above, although it's not unique to quantum processes -- but if consciousness were caused by something quantum, then you would indeed expect that two systems which could no longer entangle could no longer form the same consciousness. But this is also true for certain classical systems, like computer networks, and might just as easily be described as a network effect -- which is why I didn't include it.

Re your discussion that cars are a bad example: I agree; I used them because (1) they were your example and (2) I never proposed that we "need to bring in quantum mechanics to explain why we don't understand the brain." Again, the question I was answering was, "why are some researchers so quick to blame our failure to understand consciousness on quantum mechanics when we don't yet understand the classical parts of the brain?" -- which presumes an agreement that yes, the brain is indeed something whose classical parts we don't yet understand. The question was "why might some people reasonably think that quantum mechanics will form an essential part of the theory?"

(5) It doesn't matter what people agree or disagree. What is pivotal about the Star Trek transporter case, what I am trying to articulate, is that it creates all sorts of difficult problems about identity. In particular, quantum consciousness theories, while often functionalist and materialist, see an opposition from much of what passes presently for materialist functionalism. This particular brand of materialist functionalism has certain philosophical problems, because they conceptualize consciousness as some sort of logical pattern or arrangement. So, for example, if a computer simulated a brain, even though we would say that there is no "brain" per se, presumably these people would say that there is nonetheless consciousness.

Now, the problems are diverse, and since I mentioned above the ontological/epistemic subjectivity distinction of Searle, I would like to raise that in the same articles Searle also mentions an observer-relative vs. observer-independent distinction: whether something is a computation depends on how you look at it, he points out -- but whether you are conscious should not depend on how I look at you.

That's not the Star Trek problem. The Star Trek problem comes instead, for me, from a book called Reasons and Persons by Derek Parfit -- it's relegated to part three of the book, but otherwise it's a good read, and you can skip to the chapter if you want without missing anything. There are an abundance of things which you might think "make you you," or more generally, make you the same person as you were five minutes ago, or even five years ago. But thinking about the Star Trek transporter, and one's visceral reactions to certain stories about imperfect transporters, tends to make many of these philosophically problematic. The problem is that even though there are two people who are qualitatively identical in the world, they are not "the same person" in the above sense. So the fact that you are "the same person" as someone five minutes ago would appear to not depend on any particular quality or aggregate of qualities. Qualities wouldn't make you who you are.

You've added, as I see it, that since the universe is quantum, even physical continuity can be made epistemically ill-defined if the replicas are sufficiently exact, but I don't think materialists would have a real problem with that.

But here's where a quantum approach could have a benefit: in a quantum theory of consciousness, qualities perhaps could make you who you are. The whole thought experiment crumbles if the device is not even in principle realizable, and that's precisely what quantum mechanics purports to offer.

(5b) The "it" is quantum mechanics.

(5c) First off that's not what I claimed -- I claimed that "some of the building-blocks of warm-and-fuzzy consciousness just fall out." Second, you're in luck: if Newtonian mechanics is sufficient to you, then Ehrenfest's theorem makes consciousness fall out of quantum mechanics, problem solved. Unfortunately, your "theory" requires modification because the "thermal noise" is quantum and the "deterministic parts" are not deterministic at bottom, but other than that, I invite you to use whatever your theory is, and apply your Hamiltonian with the Schrodinger equation rather than with Hamilton's equations of motion. On average, you'll get consciousness out again. QM does not take away; it adds.


Satire, right?


Useless comment, right? What part looks like satire to you? They are all pretty standard philosophical arguments, re Searl, Penrose et al.


It's confused bullshit. The rest doesn't matter.


Citation needed.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: