## Wednesday, January 16, 2013 ... /////

### Quantum physics doesn't depend on definitions of observers

Lots of people who are trying to understand quantum mechanics but who don't really want to listen constantly ask the same question:

What is an observer really?
This question is usually encapsulated in the linguistic mud that is equivalent to the following monologue:
I'm sick and tired of explanations of quantum mechanics because they never tell me who is an observer and who isn't. Now, I am the savior of physics who will ask you and you will finally tell me and everyone else what are the ultimate, exact, well-defined criteria that determine who is an observer and who isn't, when the sound of a falling tree was heard and when it wasn't. This will permanently eliminate all the confusion about quantum mechanics. Amen.
These people must believe what they're saying but if they were also able to think about it, they would realize how stupid the question is. What kind of an answer are they envisioning if they really want to divide the objects or physical systems in the world to observers and non-observers?

Maybe, they expect you to say "An observer is someone with a social security card that must carry the signature of Barack Obama." Perhaps, an observer divides people to castes and only some of them are composed of observers. Maybe, the definition denies Darwin's evolution theory and it declares humans as observers while all other animals and organisms are qualitatively different. Maybe the humans have souls and consciousness or blessing from God and they make a qualitative difference. Maybe there's a sharp boundary between conscious processes and unconscious processes.

Needless to say, all such boundaries would be totally preposterous and their existence would violate the basic character of the laws of Nature. There can't be any sharp boundaries between observers and non-observers. No stamps or ID cards play a role in the laws of physics. There's no fundamentally qualitative difference between different species of organisms. In fact, there's no sharp boundary between microscopic and macroscopic objects. All these characteristics are continuous, gradual, and for an object to become a human is a long journey (it's easier to be born as one). And the human isn't necessarily the most perfect intelligent being that is allowed by the laws of Nature.

Most importantly, the vague and colloquial understanding of the word "observer" is perfectly enough because the actual rules and laws of physics don't depend on any details of the definition of the word "observer".

So as plain English indicates, an observer is someone (a physical system, but usually one resembling an intelligent animal or an AI-like computer) who observes something (I say "who" and not "that" exactly because that's the word we start to consider more apt than "that" once the physical systems become able to observe things and do related things!). That's everything you need to know!

There isn't any "deeper or more accurate definition" that would be needed to specify how Nature works. In physics, we use the word "observer" to describe a physical system or "agent" that is able to perceive the information about some observables (time-dependent dynamical variables that describe the state of the physical systems) and, if possible, process them. Most often, we want the observers to be able to remember the information, send it somewhere, and/or verify the laws of physics that claim to say something about the patterns relating different observations.

An observer of a particular observable, for example the number $N$ of photons in a box $\dd V$, is simply someone for whom the proposition (equation) $N=n_i$ has (or will have) a well-defined truth value. That's everything I need. The observable has a well-defined truth value because – and I hope you won't be surprised – the observer has observed the observable. ;-)

Now, to determine whether a macroscopic collection of organic molecules (a candidate animal) or a bar with semiconductor molecules (a candidate computer) is actually able to see the light (or something else) and/or remember its properties and/or calculate with it and/or use the information to optimize its behavior to achieve a certain goal, you need some specific complex disciplines of greater physics such as neuroscience or electronics or information technology.

But this is clearly not the issue that the people are asking about. I think that all of us – and most of them – understand that the neuroscience and electronics and information technologies in the real world simply aren't fundamental. Organisms and computers are complex bound states composed of many molecules that are described by the laws of quantum mechanics. I think that everyone who has at least started to think about the equations of quantum mechanics (even if incorrectly) does believe that these universal laws ultimately determine the behavior of semiconductors and proteins, too.

Instead, the goal of their question is more metaphysical or philosophical in character. They're really asking about the existence of "consciousness" because they believe that consciousness or something similar that requires a "soul" is needed to define the laws of physics. But it ain't so.

Consciousness: a great mystery whose pure part is outside science

You know, consciousness is fascinating. As a kid, I would be intensely attracted by its detachment from all empirical observations. I have consciousness, self-awareness, but you don't have to believe me because the "pure consciousness" doesn't have any visible consequences for the external observers. Of course, the "applied consciousness" does seem to have consequences. I am talking about things including consciousness itself, sometimes emotionally, I am able to say "I am aware of myself", and because you probably also have consciousness that manifests itself by talking and vibrating with your head and eyes etc., you decide that I am qualitatively analogous to you so I must have consciousness as well.

Maybe I have the same conscious feelings when I see the red color as your feelings when you see blue and vice versa (which is possible especially if you believe in some wrong political ideology). But you may dismiss these differences by Occam's razor. If the physical careers of our senses are analogous, and if we have analogous molecules in analogous cells of the retina, we arguably have the "same feelings" if a red photon hits our retina. More speculatively, we may think about "conscious feelings" that humans are incapable of. What does an uranium nucleus "feel" when it decays into decay products? This question is less urgent because the uranium atom can't process these feelings or information much so "who cares" (discrimination). But for humans, the question seems pressing: Do I have any consciousness at all?

A mysterious question, indeed. Over the years, I have lost much of my interest in this metaphysical question because of three reasons. First, almost by definition, it seems impossible to answer it within science. If I define the "pure consciousness" as something that is totally isolated and disconnected from all observations, there can't exist any manipulation with the empirical facts that would answer questions about consciousness. Second, because I don't have direct evidence of other people's (or objects') consciousness, it doesn't make much sense to study it. Third, there seems to be no sensible reason to expect that claims about "pure consciousness" may be sharp and rigorous. They are intrinsically vague for a simple reason: they have nothing to do with the quantitative things one may observe – and properties of "soul" vaguely attached to matter that can't be measured don't have any reason to carry well-defined values or rigorous laws relating these values.

So when it came to consciousness, I started to realize that only "applied consciousness" (the actual manifestations of the fact that a physical system is able to measure, remember, and process information in a sufficiently complex manner) belong to science. I would still agree that some "consciousness stripped of the dull material trivialities such as eyes, brains, and microprocessors" exists in some sense but I still find it important to appreciate that the hypothetical existence of this "conscious soul" does depend on the material carrier that may be studied by the scientific method i.e. by a careful analysis of observations. Even though I feel that "some mystery of pure consciousness remains unresolved" by science, I have also become extremely certain that "thinking about these things will never bring and can never bring anything more constructive than metaphysical flapdoodle." In this sense, I have abandoned much of my interest in consciousness for pragmatic reasons.

Maybe you are upset by the large number of paragraphs about pure philosophy – unlimited babbling. So let me get back to quantum mechanics a little bit. My main point is that none of the hypothetical claims about the existence of conscious souls influences any laws of quantum mechanics. This is a widely misunderstood point which is why it may be a good idea to mention how people like to misunderstand it.

They usually think – because they are often told – that a conscious observer causes a "collapse" which allows the superposition states $\ket\psi = a\ket{\psi_1}+b\ket{\psi_2}$ to shrink to one of the options, either $\ket{\psi_1}$ or $\ket{\psi_2}$, with probabilities given by $|a|^2$ and $|b|^2$, respectively. When this collapse happens, someone is perceiving that something has happened, something has been measured. Because this collapse is such an important intervention into otherwise smooth and regular laws of evolution of the state vector according to Schrödinger's equation, evolution that is supported by all the evidence, one should need some special "stamp" – such as the social security card with Obama's signature that I started with – to interrupt the peaceful Schrödinger's evolution and to replace it by the "collapse".

I am writing down this preposterous story because this is exactly the type of thinking that many popular – and, using Sidney Coleman's words, sometimes even not-so-popular – books and articles want you to manipulate you into. GRW and Penrose collapse theories as well as the many-worlds ideology are example models giving special objects the right to "intervene" into Schrödinger's equation, either by discontinuous jumps or collapses or by splitting the world (which is comparably, infinitely ambitious). However, all this reasoning is completely nonsensical. There doesn't exist any systems for which the evolution according to the laws of quantum mechanics such as Schrödinger's equation is replaced by some discontinuous jumps. Quantum mechanics applies to all systems and processes in Nature, regardless of their size, duration, sex, race, and nationality.

Instead, what an observer does when it "measures" the value of an observable such as $N$, the number of photons in a region, is that it simply attaches a value to $N$. Equivalently, it ascribes the truth value to all propositions of the form $N=x$ or $N\gt x$. Who has the right to do it?

The key point is that to define the laws of physics, we don't need any definition or criterion here. Why? Because when an observer ascribes a value to the observable $N$, it has absolutely no impact on other observers' description of the reality! Indeed, as the Wigner's friend thought experiment was designed to explain, other observers – if they want to make really accurate predictions about their future observations – should keep on treating the "conscious observer" as a dull physical system that evolves into the state vector that is a general complex linear superposition of eigenstates with different values of $N$. In practice, one may use a description in which the truth values and/or probability distribution become "classical" but all these descriptions are at most approximately valid.

So whether the "conscious observer" has "perceived" the value of $N$ makes absolutely no impact. I still need to describe the degree of freedom $N$ in terms of probabilistic distributions – and indeed, in quantum mechanics, I need interference-capable complex probability amplitudes. I only reduce my probabilistic reasoning to a fact-based one once I learn about the value of $N$ or something else myself. At that moment, I ascribe a value to $N$ or another observable (which, equivalently, changes the state vector or density matrix I am using to describe Nature – these objects represent the state of my knowledge). I ascribe the truth values to various propositions. Whether a different observer ascribed a value to a proposition about an "intermediate question" makes absolutely no impact on my predictions – and my predictions are by definition the only physically "existing" knowledge about the physical system that I have.

It is wrong for me to ascribe particular values to observables if I don't know their values. It is wrong for me to ascribe particular truth values to propositions whose answers are unknown to me. That's why another observer's act of ascribing a value to $N$ just makes no impact on my knowledge – only if I learn about $N$ myself, it influences my predictions!

I have already mentioned that whether a physical object may actually see some information or calculate with it is a question for neuroscience or electronics or information technologies. I have already mentioned this point. But you expect some restrictions. Certain things can't be "perceived" at all. Indeed. When you ascribe truth values to some propositions, the operators expressing these propositions must behave as if they were classical numbers $0,1$ able to be added and multiplied according to the classical rules. For example, if you study which of the alternative histories will occur, the alternative histories in your set must be consistent histories. In practice, it means that unless these histories are artificially engineered and very contrived, they must be histories talking about the values of some "quasiclassical" observables according to some classical limit of your quantum theory.

Nevertheless, I could even extend the definition of an observer a little bit and allow him or her or it or them to "observe" the truth values of propositions that aren't logically compatible in the sense of "consistent histories". Such an observer would be an illogical observer or a confused observer. ;-) It often looks like most people fall into this generalized category of observers. :-) In the same way, there may be sloppy and inaccurate observers – those whose observations are sloppy or inaccurate. More seriously, the inconsistencies between the observations by the confused observers would be analogous to the "paradoxes" that appear when you try to interpret GHZM and similar quantum games classically.

Fine. If you want an observer who sees and perceives real facts – sufficiently accurately – and who processes them, you need an observer that has well-functioning eyes (or an equivalent measuring apparatus), a brain including the memory (or its electronic or another replacement), and so on, according to the rules of neuroscience or electronics or information technologies or something else, and this observer must work with questions and alternatives that are logically consistent in the same sense as "consistent histories".

However, I still need to emphasize the main point of this article. All these features of a good enough observer may or may not be imposed – and it makes exactly zero difference for everyone else! You only impose the "consistency of histories" because you want to be an "unconfused observer". At the end, the only special (or additional) feature of an observer is that he or she or it or they observe something. And observing something isn't a crime. For the observer to have a consistent logical framework with truth values of various propositions about observables (for his histories to be consistent), this observer will have to send actual photons somewhere that perturb the system they're observing (or intervene in an analogous way).

But this necessary perturbation is a mechanical rather than metaphysical process. If some photons hit electrons and if you're another observer in the room, you need to calculate with quantum mechanics for several particles and you get different predictions for the pattern drawn by the electron – effectively, the interference pattern is destroyed because of the electron's entanglement with the escaping photon. But you don't have to "know" whether someone in the vicinity of the photon was an observer or conscious or human or animate or anything like that. It's the photon hitting the electron that destroys the interference pattern, according to the standard rules of quantum mechanics, not the soul or other mysterious anthropomorphic features of the surrounding physical systems claiming to be human!

Once again, the only special or additional properties of observers by which they "exceed" the generic physical objects around them is that they observe. By observing things, they ascribe values to some observables. And the laws of quantum physics imply probabilistic relationships between these values at different moments. An observer who is really worth the name may verify these relationships. That's it. But other observers such as yourself ascribe values to other sets of observables and all the other observers may always be treated as dull, unconscious objects! Only if two observers – two physical objects "personifying" two sets of consistent histories – include the same questions/propositions into their logic (scheme that assigns the truth values to propositions about observables), quantum physics guarantees that the values will agree.

Needless to say, people are looking for an "objective classical model of reality" that is valid for everyone. But quantum mechanics shows that Nature can't be described in this way. Instead, quantum mechanics tells you that you must understand yourself as an observer who may perceive the values of certain observables and quantum mechanics tells you that observing some values of observables at one moment implies that the probability of observing some combination of other observables at a different moment is something or something else. That's the only thing you may really empirically verify so it's just unphysical to "demand" that science also explains something else (such as an "underlying objective reality"). And indeed, it's very important in quantum mechanics that physics can't objectively describe too many things, that it is always impossible to ascribe values to too many observables or truth values to too many propositions. The uncertainty principle is just one of the basic formulations of this fact: you can never ascribe exact values to two complementary variables such as the position and the momentum of the same object. But this principle is really omnipresent and essential in all of quantum mechanics and may be formulated in many related ways, seen in many reincarnations. You just can't talk about the objective properties of (most of the) things you can't observe.

To summarize, an observer is someone who may observe things and verify the predictions from the laws of physics, among related things (e.g. using them to his advantage). But if there's no one who can do it, no one is hurt! Nature may seem like it has no purpose – but it's how Nature is according to science, anyway. What's important is that the "consciousness" or "act of pure observation" or "realization" doesn't make absolutely any impact on physical predictions that other observers may prepare. For other observers, the original observer is just a piece of matter, a dull physical object. That's why, if they're sensible and physically oriented, they don't spend hours by trying to find an exact definition of an observer. They know damn well that they're able to ascribe truth values to propositions about observables and that's enough for them to verify the laws of physics. One can't expect any sharp rule dividing physical systems into castes, into observers and non-observers. No such sharp rule exists and no such sharp rule is needed in physics.

And that's the memo.

#### snail feedback (41) :

Dear Lubos,

I would say that the observer is simply a macroscopic quantum system, which interacts with my original microscopic system. The interaction leads to decoherence, so one can effectively replace the wavefunction of the system with classical probabilities.

Consciousness clearly doesn't matter, the interaction with an arbitrary environment always leads to decoherence.

The air molecules in the box of the cat cause the same effect that a conscious observer would cause.

Only what we observe produces our conscience.

Dear Lubos, A question from the peanut gallery: in the two-slit experiment I seem to recall -- in Susskind's video lectures I think -- that if an electron interacts with the material in which the two holes exist, on the side of one of the holes let us say, in such a way as to leave a visible mark or piece of evidence behind that such an interaction occurred, that this would count as an observation and the interference pattern would disappear. Is this correct?

What would happen if you begin with two holes which are precisely of the same diameter but gradually made one of them smaller and smaller until it finally disappears. Would the interference pattern gradually be replaced by a non-interference pattern?

Can an observation be described as any macroscopic effect which a human observer can in principle verify either at the time it occurs or later, it making no difference when?

I hope these are not idiotic questions.

Dear Luke, nope, they're totally sensible questions and your definition of an observation is much better than what one usually sees.

If two slits in the double-slit interference have different size, it indeed means that the interference partly disappears, there won't be any "perfect minima" where the particle can't land.

It's because one is simply summing

A * exp(i*phase1) + B * exp(i*phase2)

where the absolute values of A,B are different because the slits have different sizes. If A is greater, then the slit proportional to A dominates and one may see something close to the picture one gets from the slit A only. In such a picture, the interference pattern is subdued - because terms of order A* B are smaller than A* A.

Lubos,
What do you think of this course by Spekkens on the foundations of quantum mechanics? At Quantum Frontiers everyone called it excellent.

And what do you think of the great (or not?) accomplishments of the Quantum Foundations Group at Perimeter Institute? I wish you would write a separate post discussing the merits or lack thereof of their much touted work.

I have my own opinions about the matter, but would like to hear yours.

http://perimeterscholars.org/301.html

http://www.perimeterinstitute.ca/people/research-area/quantum-foundations

Do you mean consciousness? ;-)

Man thats a lot of words. I almost tl;dr-ed it. I'm only an amateur but isn't decoherence the best way to understand the nature of an observer? Any interaction counts as an observation in proportion to its ability to cause decoherence.

As for the consciousness issue... you seem to be suggesting that since there is no way to tell the difference between a philosophical zombie and a conscious person then the question isn't really a science question. If consciousness has no observable effect then all it does is make us helpless observers to events that we have no control over.

Yet if consciousness has no physical effect then how can we talk about it? If consciousness did not exist would the philosophical zombies that evolved in our place talk about the consciousness that they do not have as if they did?

This ability to discuss consciousness is arguably a physical and observable consequence of consciousness.

Dear Lubos,
as a long term reader of this blog I know your view on quantum mechanics and I would say it has deepened my understanding of it. But I think we never discussed about consciousness. I would say there is actually a way to check its presence empirically and it is called the Touring test. It should test for intelligent behaviour but I would claim this goes together with consciousness. The more advanced the computer programs become the more I would support Penrose that computer programs as we define them today can never show true intelligence or become conscious, Even for the hard to beat chess programs it is easy to create chess problems which can be solved by every bright child but not by them. So I find it a reasonable speculation that the construction of an intelligent computer may require to make use of the laws of quantum mechanics. By the way I found that even made a remark that he finds this view of Penrose a reasonable speculation. I would also tend to think that consciousness is a kind of independent actor which may be influenced by feelings of hurt etc. For a computer you may naively try to model the degree of hurt it "feels" by a counter but why would this counter need to be accompanied by an actual feeling, The physical laws also happen without feelings.

Classically, you cannot get complete destructive interference (cancellation) of the two waves unless they have exactly the same amplitude, i.e. the two slits transmit the same amount of light. Since the quantum mechanical description has to eventually reduce to the classical description no spot on the observed screen can have zero probability of being hit by a photon unless the slits have the same width.

Since I wasn't so stupid on my last question, let me try again. Imagine me as a male version of Penny across the hall:

When I look at the wingback chair in my library am I staring at the average pattern of quantum events (of electrons interacting with photons) of countless billions of atoms arranged into atoms arranged into molecules in the shape of a chair? And if I am, then isn't a macroscopic object in a sense a pattern of events as described by the equations of quantum mechanics. Only the pattern is real.

You can give me a Sheldon look now.

I predict (before Lumo wakes up!) that you will remain on good terms with Sheldon even after this last sticking-your-neck-out question. ;)

IMO, the word "real" can be etymologically+philosophically tenuously extrapolated to a fuzzy notion of 'ultimate Reality' as an Eternal (but not classically 'timely') Patterning (from a Platonic blue-print of infinite possibilities & impossibilities) Tendency. %-}

Oh what a satisfying and super-witty (except for the slightly worn-out reference to a certain birth-certificate) sorting out of this issue this article was!!!

You Lumo deserve - do according to my naturally limited (but by definition always percEPTive) observations and estimations of value - a major reward for being/writing the ways that you are!!!
--
P.S. It should be obvious that I am emoting.
In practical reality, the size of this proposed (and wished for) "reward for TRFicness" will of course be determined by available financial means and by the degree of persistence of relevant positive motivating influences from several different, including unrelated, sources.

Can I kindly ask commenters what they think about brain processes? After all, brain is supposed to be the physical system giving rise to consciousness. Do you think that consciousness, thought and cognitive processes are fundamentally based on classical physics (like computers) and therefore deterministic in nature or not?

Good stuff. Sometimes I think that, even though like you I started with wondering about consciousness and physics at some point, I have been saved from fluff because of all those bubble chamber pictures I scanned in my youth. Film is a good observer :).

Consciousness is an ill defined term. Observation is on other hand a poor choice of word for a well defined term. Observation of possible outcomes is only by a consistent history constrained. I think I am observing this blog entry and all its content. but if I were a particle affected in a casually important way I would be just as much an observer and observed by the interaction that localized me in a causally important way.
If that is what you are expressing then I exist and are right. Otherwise be kind in disputing me as I may be part of your consciousness and you not I am wrong. :-) I always preferred Ambrose Bierce "I think therefore I think I am" to Descarte. Perhaps the former President Clinton should be given an honorable mention with it "depends on the meaning of is". Anyway philosophy is fascinating too so at least this reader enjoyed greatly what you called blather.

It seems to me that a separable QM system with some chosen basis should qualify as a generic observer.

Dear Mikael, I may share your beliefs - consciousness goes with the manifestations and may be measured by the Turing test. Fine. But what if not? What if the machines are really always "dull and dead", not aware of themselves, however accurately they emulate the human behavior? While this possibility may look contrived, it seems impossible to become certain that it's invalid.

Maybe, I don't get the difference... "la conscience" in French means both "conscience" and "consciousness" in English...

I see, that doesn't make it easier. ;-) The difference is easy to understand, though.

"Consciousness is the quality or state of being aware of an external object or something within oneself."
http://en.wikipedia.org/wiki/Consciousness

"Conscience is an aptitude, faculty, intuition or judgment of the intellect that distinguishes right from wrong."
http://en.wikipedia.org/wiki/Conscience

What do you mean by separable? If you mean that it's described by a small Hilbert space such that the total Hilbert space is the tensor product of this Hilbert space and another one for the remainder, this condition doesn't hold in quantum gravity and isn't really necessary for anything that we demand observers to do.

If you mean a related thing that the observer is a subsystem that is unentangled with others, it's way too constraining because observers surely do get entangled with the rest of the world - this is really necessary for decoherence which is needed for them to observe the facts as classical facts.

A chosen basis... One may choose a basis at all time but any choice of basis doesn't really correspond to an observation - an observation is usually much less refined and it is so for fundamental reasons. Moreover, one should only be observing a subsystem, i.e. choose a basis for a tensor factor, but this factorization of the whole Hilbert space doesn't hold exactly, as discussed above.

So your definition really can't hold for tons of reasons. But my real complaint is that it has no meaningful purpose. Why would you introduce the term "observer" for the arbitrary set of conditions that you have outlined? In physics, we use the term "observer" for a totally different reason that actually agrees with the linguistic origin of the world. It's a physical system that may observe, i.e. ascribe values to observables. We use this concept because the observer may verify the laws of physics and do related things. If he can't do it, he isn't really the observer we're looking for. By trying to associate some objective mathematical properties to an observer, you're really missing the point of the concept "observer" - why it's used by physicists.

Thanks, Peter, for your excitement! ;-)

Dear Paul, the ability to cause decoherence is proportional to observation... I don't know what to do with such things. Decoherence is a genuine emergent effect in QM, a description why classical truth values of propositions emerge from the originally interfering, different quantum probability amplitudes. But why should one identify decoherence with observations? What do you want to achieve by that? We're using the word "observation" because someone may actually perceive and process information.

Burning coal experiences lots of decoherence - which may surely be applied to this system - but I wouldn't say that burning coal is a particularly good or strong example of an observer because the coal is just too dull and stupid.

Yet if consciousness has no physical effect then how can we talk about it?

Why should we be unable to talk about it? A tape recorder doesn't have consciousness, at least not the human-sized one, but it may still talk about consciousness in the same way as we do if you record such a monologue. So talking about consciousness is surely a proof of nothing mysterious or soul-like.

QM remains consistent to all observations regardless subjective belief and that is also fascinating. But, my question is whether you are imposing a further constraint on an observation of it being subjective in any greater sense than that it is an outcome of some interaction , ie an event causally consistent with all other possible measurements were there someone measuring? If no one had been here to measure and notice QM is a description still the laws of physics would be the same and so I am not sure why the word subjective needs be included with measurement.

Dear Robert, I understand that people often say that the "observation is subjective" is a constraint. They say it to express that they are uncomfortable.

But the reality is exactly the other way around. When you say that observations reflect something that exists even without observations, you are imposing a highly nontrivial constraint. Roughly speaking, this proposition demands physics to be classical physics.

But classical physics is a special case (hbar to 0 limit) of quantum mechanics, not the other way around! In the classical limit, things may be considered objective. The objective reality emerges with an arbitrary accuracy if hbar is arbitrarily small relatively to the quantities describing the studied physical system.

The realization that the observations do *not* have to reflect any objective reality isn't a constraint at all: it is a removal of a constraint. And indeed, the observations force us to get rid of this constraint, to jettison this excess baggage in order to make progress. The laws of physics directly link the observations with each other - probabilistically - and they just don't allow one to "insert" some objective reality in between them. Instead, using Feynman's path integral approach, all possible realities must be summed over to calculate the probability ampitudes of a transition from an initial state to a final state. But none of the intermediate histories is more real than any other. They're just terms in a complicated calculation and only the final result of the calculation makes a physical sense.

All these things - totally universal and rudimentary facts about quantum mechanics - are equivalent to the (more philosophically sounding) proposition that the observations in QM are fundamentally subjective. The word has to be included because if the observations were objective - objective processes of any kind - according to a theory, the theory would simply be conceptually a classical theory. That is really equivalent to hbar=0. But we know that hbar is not zero so the observations and other acts can't be objective changes to a physical system. They're subjective.

I believe (not fanatically or strenuously) that while "to merely and purely think" that building conscious (in the way near enough normal people are aware reflective and feeling) copies of our evolved brains (or for that matter other brainy animals' brains) is obviously one of the infinite possibilities of 'the eternal patterning tendency', to 'practically build' such things (such extraordinary psychological things) is fundamentally prevented because it is one of those things that are infinitely impossible to build WITHOUT an immensely unlikely but possible 'bio-evolutionary patterning trajectory'. %-}

Because I see it this way I have little respect for people who (some of which are referred to as AI experts) seems to take seriously the idea of building conscious computers (such as likewise IMO ultimately impossible to build multi-purpose computers that calculate with qubits).

I can talk to my telephone and sometimes be correctly understood and sensibly talked-back to by it. So much for consciousness defined as a capability of being on "talking terms".

Yes. I see that. I was merely misinterpreting your use of subjective.

I think decoherence lets us see the collapse of the wave function as a physical process and so we no longer need to talk about observers at all. It is true that you need an observer to talk about what actually happened from the point of view of that observer. But as you say QM does not depend on how we define that observer. A rock is complex enough that an interaction with it will cause decoherence. We can treat it as an observer. If the qbits of a quantum computer interact with the rock they will collapse and destroy the quantum computation in progress just like if you looked at the qbits. The fact that the rock cannot talk and probably is not conscious is irrelevant. It functions as an observer.

The tape recorder... yes in a sense it can at least seem to talk about itself. But the point is could such a tape recorder evolve?

Sure you could program a mindless zombie to talk about the consciousness that it does not have. But what evolutionary force would cause the zombies to evolve the ability to talk about consciousness as if they possessed it? If consciousness has no effect then by definition there can be no selective pressure for it or even an illusion of it. Yet at least the illusion of it exists.

I'm not trying to argue for some supernatural soul here but I do think there is still a mystery. I cannot imagine what the solution to the mystery would look like and that can be a sign that it is an illusion. But the self referential argument above makes me skeptical of that. After all what sense does it make to claim that the ability to have an illusion is itself an illusion?

Right, Paul, physics works without any concept of an observer (except for "us", each of us may call himself an observer but it plays no role) and decoherence is a part of the explanation why it does reproduce the classical intuition when it's applicable.

Your comments about evolution pressures for tape recorders are surely interesting if they work but I don't understand either of the logical links that you need to establish what you need to establish.

I don't understand why - and you seem to assume that - consciousness should have a necessary condition that it has spontaneously evolved in competition. Moreover, I don't understand in what sense tape recorders talking about consciousness haven't evolved. They were produced in a competitive environment that was evolving much like organisms and other things. At some moment, people recorded and sold CDs talking about consciousness (and especially souls and other esoteric stuff) because these occult consciousness-discussing CDs or tapes (sorry, I will use it interchangeably) had an evolutionary advantage in the natural selection and competition against other types of CDs such as the sound of squeaking doors or lectures on physics.

Thanks Tobias. Indeed I meant consciousness. ;-)

Lubos,

Yes tape recordings of talk about consciousness "evolved" in a sense. But they evolved in an environment where there were people claiming to be conscious. Could they have evolved in an environment that had no people claiming to be conscious? Clearly not.

People evolved and claim to be conscious. What environmental condition caused this? If we removed that condition would we have evolved a technological civilization of people who make no claim to being conscious?

The consciousness talk on the tape recorder is contingent on the consciousness talk of people. What is the consciousness talk of people contingent on?

Dear Lubos, I agree that it is not a scientific question to ask whether a machine or other human being experiences consciousness, Maybe if somebody had a theory of consciousness and could link it to something we could measure in the lab the gap could become very small.

In any case I find it a very interesting scientific question to ask whether we can ever build an intelligent machine and what are we missing right now so that we can't do it. You dismissed the possibility that quantum mechanics plays any role. But do you think that an intelligent enough programmer could write a program which passes the Turing test and why are todays programmers failing?

Sure it's possible for a species of zombies to evolve a delusional belief in consciousness, and talk about it as a side effect of the evolution of language and linguistic metaphors.

So, if you have a social species with language, and members of that species evolve the ability to answer questions to others about their internal states, and they have short term memory, and they subsequently internalize such an ability as "private thoughts", and their language has first person pronouns and a symbolic metaphorical concept of a "self", these members can easily delude themselves into thinking they have nonexistent consciousness.

Maybe the human species is such a species, and all humans are zombies, including you and me and Lubos ;-)

Empirically, it's not possible to tell if Lubos is a zombie or has a soul ;-) So, it makes no scientific sense to talk about Lubos' consciousness ;-) ;-) ;-)

Maybe Lubos **really** is in some quantum superposition from "our", but not "his" subjective viewpoint ;-)

> I would still agree that some "consciousness stripped of the dull material trivialities such as eyes, brains, and microprocessors" exists in some sense

> I feel that "some mystery of pure consciousness remains unresolved" by science

You surprise me. Are you not a materialist then? ;-) You're also deviating from your own party line on the positivistic definition of existence. **gasp**!!! Do you have, **gasp**, faith?

You're right. There's no "natural" or "simple" division of physical objects within the universe into observers and non-observers. There's also nothing special about human brains.

That doesn't rule out the possibility the division line lies outside the universe, though.

You say quantum physics doesn't need an observer, but without an observer, everything will remain in an uncollapsed superposition, and you get MWI. But you reject MWI.

If everything is subjective, surely a subject is needed?

According to consistent histories applied to the universal wavefunction, the probability that we exist is exponentially small.

What?! You heard me right. There were quantum fluctuations of order 10^-5 during the inflationary epoch, but they were quantum. During reheating and after, this superposition remained, albeit in decohered form. The details of the structure formation which followed depends on these CMBR anisotropy quantum fluctuations.

The evolution of life here on earth depended on mutations, and most mutations are caused by quantum events like cosmic rays, radioactivity, or quantum chemical interactions between carcinogens and the DNA. Schrodinger's cat also applies to DNA mutations, you know. Most chain of mutations do not lead to the evolution of humans.

Schrodinger's cat applied to bubble chambers.

Here goes. A charged particle in a superposition of positions passed through the cloud chamber. It entangled with water molecules in the chamber leading to bubbles condensing. Photons streamed through the chamber and some of them got entangled with these bubbles. Some of these photons got entangled with the film. The film got entangled with photons in the room, and some of these photons got entangled with your eyes and brain. Oh my god!!! You were in a quantum superposition!!!

Dear KPM, I asked you to reduce the number of your crackpot comments a week ago, you didn't, so I just banned you and all your ways to identify yourself.

An observer is simply anything at all even non-conscious and specially non-conscious that affects the sub-atomic particle that is being observed: because in the
sub-atomic world or environment everything is connected to everything else, in particular when the connection is close enough as to affect the observable which is another material thing, like yes a particle.

Now, the big difference between a non-conscious observer like
another particle that is close enough to influence or affect directly and immediately the 'observable' particle, and a conscious observer say like a human being, is that the conscious human can decide to not be present in the immediate and direct environment of the 'observable' particle, by going away.

That option to go away is due to with humans the power of free
will.

Now, from this explanation it is obvious that the 'observer' is
also influenced or affected directly and immediately by the 'observable', since they are both correlation-ly 'observer' and 'observable' in their common sub-atomic environment.

In the case of a human observer, he is linked to the observable particle by way of instrumentation of detection used by him to observe the observable particle, and thus the physical instrumentation sends particles to the immediate and direct environment of the observable particle, which then influence i.e. affect directly and immediately the observable and now observed
particle.

Marius de Jess