Is Consciousness involved in wave function collapse?

[Varon]

Another thing. Before humans existed on the planet. We had plants, sea creatures, dinosaurs, birds, chimpanzees. Now for evolution to occur at all, the biochemical parts have to be in definite state and outcome. This is because things in mixed state can't evolve into complex things. Therefore the fact that evolution can occur without humans prove that definite outcome occurred in the particles and chemical systems enough to evolve and produce humans. Therefore, human consciousness is not sufficient to to produce definite outcome. I know the arguments that quantum theory only produce ensembles.. why individual outcome occurs is due to humans as the saying goes. But in evolution before humans, the particles and chemical system have individual outcome because mixed state don't offer many degrees of freedom.

 

[Ken G]

Earlier you said that machines as measuring tool can only produce mixed state..

Actually, what I said is that machines would never need to develop a concept of anything but a mixed state. A "mixed state" is not a real thing, it is an element of a theory.

this meaning the full system is still in pure state.

Yes, that is how the theory of quantum mechanics would frame the situation.

And you said it is only human that cause the single outcome.

I'm saying that it is only our experience of reality that causes us to find any fault with the description of formal quantum mechanics mathematics, yes.

Are you claiming that human consciousness can produce the born rule? Then this is the same as saying that consciousness can collapse the wave function.. because the born rule is synonym to wave function collapse.

I have no idea "what collapses the wave function" (in the sense of the single outcome, not the decoherence), indeed I believe, as I said, that this question will never be answered by science, because it is too deeply embroiled in our own role in our science. That's a very "CI" perspective. What I said about consciousness is that its presence is what we can trace the very issue to in the first place.

 

[Varon]

Actually, what I said is that machines would never need to develop a concept of anything but a mixed state. A "mixed state" is not a real thing, it is an element of a theory.
Yes, that is how the theory of quantum mechanics would frame the situation.
I'm saying that it is only our experience of reality that causes us to find any fault with the description of formal quantum mechanics mathematics, yes.
I have no idea "what collapses the wave function" (in the sense of the single outcome, not the decoherence), indeed I believe, as I said, that this question will never be answered by science, because it is too deeply embroiled in our own role in our science. That's a very "CI" perspective. What I said about consciousness is that its presence is what we can trace the very issue to in the first place.

What? You kept saying that consciousness is what produce definite outcome from the ensemble in the mathematics of quantum theory. Now since definite outcome can only result from collapse of the wave function. Therefore from the logic, it goes that consciousness is what collapses the wave function.

 

[Ken G]

Ken G - we both have different opinions on what makes a beautiful theory. This is why our views differ.

Well, the axiomatic structure of quantum mechanics is widely viewed as one of the most beautiful in all of science, and indeed this is pretty much singlehandedly the motivation for string theory and the search for a graviton model. But I will grant you that just what entails "beauty" is subjective. All the same, my point holds-- if all we had were mindless robots, they would never have any reason to even dream up the concept of a single outcome of an experiment. They could be built to test, process, and solve the equations of formal quantum mechanics, pass every test with flying colors, achieve survival advantages, all without ever imagining that there is any such thing as a "single outcome" to a decoherence episode. It must have something to do with the necessity of modeling itself as part of "what is happening." I don't know exactly what point we would call it "consciousness", that's why I didn't try to parse the differences between consciousness, intelligence, and perception.

 

[BruceW]

I feel my view is less committal, in that I don't suppose that the consciousness of a being decides whether it can or can't cause a non unitary process to occur.

 

[Ken G]

But you still haven't answered my question from above. I feel answering that question is crucial: why would a lifeless robot that did not model itself as part of its processing effort, so that did not build into the analysis a sense of "its own unique experience", ever have a reason to concoct a concept of a single outcome of a decoherence episode, where by decoherence episode I mean in the von Neumann sense of coupling a quantum system to a macro pointer to bring the quantum system into the robotic analysis?

Imagine you are programming a computer to solve a quantum mechanics problem. You evolve the wavefunction, and model whatever decoherence occurs in that evolution. You're done-- why would you ever include a final step where you choose a random number and assert "this is what happened", if there is no perception of what happened? The program could test its predictions with an apparatus that also generated nothing but mixed states, as quantum mechanics predicts it would.

 

[Varon]

Ken, I wonder how to use the concept of "Preferred Basis" in the debate. I know that preferred basis is chosen by the measuring device. Are you saying that without human consciousness the device can't choose one set of values of the basis on its own? Let's make an actual example so we can understand what you mean that only human consciousness can produce definite outcome.

Let's say the human civilization left a double slit device in pluto before the whole civilization perished. The double slit now runs on its own from either solar or nuclear fuel. The detector in the doublet slit is enough to define the preferred basis and even to produce definite values or outcome right? Or else no photons or electrons won't be detected in the screen.

Now your argument is that without a human seeing the detector values. The facts can't be known. But let's say a lifeless robot or android produced prior to the demise of the human race goes to pluto to retrieve the double slit device and read the detector. Then it can know the definite outcome. All this without the occurence of humans which are long dead.

How do you use Copenhagen to argue that the lifeless robot won't read the screen definite output?

 

[Fredrik]

Can I ask something? Does a wave function collapse upon interaction with anything? Like when a photon strikes an electron out in outer space and ionizes an atom.

I'm not a fan of the term "collapse" because it can mean different things. It can refer to an interaction that sends the off-diagonal components of a density matrix to exactly 0 instead of just making them really small. (This would contradict QM, but it seems to be what most people mean). It can refer to an interaction that makes those off-diagonal components really small. (This is decoherence). It can even refer to the selection of a subensemble on which we intend to perform experiments.

Assuming that we're talking about what interactions that should be thought of as "measurements" (i.e. collapse=decoherence), there is no fundamental difference between measurements and other interactions, but there is a practical difference. Consider a Stern-Gerlach experiment. A silver atom is sent through an inhomogeneous magnetic field. The atom first interacts with the field, and the interaction correlates spin states with momentum states. Then the atom interacts with a photographic plate. This interaction leaves behind a record of what just happened in the form of a tiny dot on the plate. That dot is the reason why this interaction is considered a measurement, while the first one isn't. (The position of the dot is interpreted as the result of a momentum measurement, which indirectly tells us the value of the spin). But even an interaction that leaves persistent records behind can in principle be reversed. (This reversal would delete all the records). That's why I'm saying that there's only a practical difference between measurements and other interactions, and not a fundamental one.

In principle, you could have an excellent reason to think that the measurement has been performed (i.e. that collapse has occurred), moments before everything that has happened to a larger system that includes you is reversed. Obviously, this would delete all your memories of ever having performed the experiment. Don't make the mistake of thinking that you could just ask your friend what the result was, because if he has a record of the result, in his brain or elsewhere, then it only means that the interaction that we considered a measurement hasn't been completely reversed.

Did you delete the question while I was answering it? I can't find the post I quoted.

 

[Ken G]

Ken, I wonder how to use the concept of "Preferred Basis" in the debate. I know that preferred basis is chosen by the measuring device. Are you saying that without human consciousness the device can't choose one set of values of the basis on its own?

No, the "preferred basis" of a measurement is something quite physical, it has to do with the type of decoherence that the measuring device creates. In our world, we say the photon made a spot on the wall, but in quantum language, the wall acted like a wall expressly because it set up the kind of decoherence that turns a photon wavefunction into a mixed state of probabilities of various spots on the wall. That's what a "detecting wall" is to a photon. So the preferred basis, here a "spot" basis, is determined by the physical interaction itself-- other bases would not see a decoherence between their eigenstates.

Let's say the human civilization left a double slit device in pluto before the whole civilization perished. The double slit now runs on its own from either solar or nuclear fuel. The detector in the doublet slit is enough to define the preferred basis and even to produce definite values or outcome right?

The detector is still physically capable of creating the decoherence that creates a mixed state for the photon. That's not a definite outcome, it's a probability distribution of definite outcomes. It's also the only predictive element of quantum mechanics.

But let's say a lifeless robot or android produced prior to the demise of the human race goes to pluto to retrieve the double slit device and read the detector. Then it can know the definite outcome.

Why do you think it can know the definite outcome? Formal quantum mechanics certainly doesn't say it can. You are basing that claim on your personal experience-- but that has to do with your own perceptions, not the action of a robot as described in quantum mechanics. So you tell a story that is consistent with your conscious experience, and you fudge the quantum mechanics theory to make it agree (standard CI practice). Nothing wrong with that, but recognize the reason you are doing it is your own conscious experience has to make sense of that physical scenario. We have no idea what the robot will experience, because robots don't experience anything.

How do you use Copenhagen to argue that the lifeless robot won't read the screen definite output?

In its purest form, Copenhagen says that physics is a tool used by a physicist, not a robot.

 

[Varon]

No, the "preferred basis" of a measurement is something quite physical, it has to do with the type of decoherence that the measuring device creates. In our world, we say the photon made a spot on the wall, but in quantum language, the wall acted like a wall expressly because it set up the kind of decoherence that turns a photon wavefunction into a mixed state of probabilities of various spots on the wall. That's what a "detecting wall" is to a photon. So the preferred basis, here a "spot" basis, is determined by the physical interaction itself-- other bases would not see a decoherence between their eigenstates.
The detector is still physically capable of creating the decoherence that creates a mixed state for the photon. That's not a definite outcome, it's a probability distribution of definite outcomes. It's also the only predictive element of quantum mechanics.

Why do you think it can know the definite outcome? Formal quantum mechanics certainly doesn't say it can. You are basing that claim on your personal experience-- but that has to do with your own perceptions, not the action of a robot as described in quantum mechanics.

Can you be clear what you meant by "definite outcome". Is it the same as one set of basis values chosen (one of the eigenvalues chosen)? If it is. Then the definite outcome are simply those hits in the detector screen which the lifeless robot can simply detect with its robotic vision.

So you tell a story that is consistent with your conscious experience, and you fudge the quantum mechanics theory to make it agree (standard CI practice). Nothing wrong with that, but recognize the reason you are doing it is your own conscious experience has to make sense of that physical scenario. We have no idea what the robot will experience, because robots don't experience anything.

In its purest form, Copenhagen says that physics is a tool used by a physicist, not a robot.

 

[Ken G]

In principle, you could have an excellent reason to think that the measurement has been performed (i.e. that collapse has occurred), moments before everything that has happened to a larger system that includes you is reversed. Obviously, this would delete all your memories of ever having performed the experiment.

This is a hugely important issue, the question of whether or not that is really possible. It boils down to how seriously we take the formal theory of quantum mechanics. In particular, do we give primacy to the mathematical structure of that theory, which says exactly what you are saying is possible (in principle, sort of the way quantum tunneling of macroscopic objects is possible in principle), and say the person's experience is subordinate to that, such that it would indeed be possible to erase all their memories and leave the universe in a noncomittal stance as to what actually happened? Or do we say that formal quantum mechanics theory is a tool of that person, not its master, and so if they experience something, they cannot unexperience it, regardless of what the theory suggests? The former approach is more in the rationalistic spirit of many-worlds kinds of thinking (it amounts to re-enmeshing the divided worlds, something we rarely imagine but is certainly possible in principle), the second approach is more in the empiricist spirit of the CI.

I take no stance on either interpretation here (though I generally prefer CI), I merely point out that the whole reason we have to ask this question is because of that moment of conscious awareness that we are loathe to imagine can get "erased." If we flip a coin, and never look at it, and then something happens that unflips the coin, we can say that whether the coin came up heads or tails is something the universe has not taken a stance on. But if we look at the coin and register heads, can something happen later that makes the universe also take no stance on what the coin came up? This is why many flinch at accepting the many-worlds interpretation, it subordinates the very experiences we are using physics to try and understand to the physics that we are using to understand them. I don't say that makes many-worlds wrong, I just notice that the cause of our flinching is our conscious experience.

 

[Ken G]

Can you be clear what you meant by "definite outcome". Is it the same as one set of basis values chosen (one of the eigenvalues chosen)?

It is not the same as the set of eigenvalues, it is one eigenvalue.

Then the definite outcome are simply those hits in the detector screen which the lifeless robot can simply detect robotic vision.

Again, it is only your conscious experience of how vision works, your own personal perception of it, that leads you to say that. If all you had was the mathematics of quantum mechanics, but no conscious experience to draw on and no self-modeling in your analysis, you would never have any reason to think that your statement above was true. It just plain does not come from the mathematics of quantum mechanics, so where does that expectation come from?

 

[Varon]

It is not the same as the set of eigenvalues, it is one eigenvalue.

Again, it is only your conscious experience of how vision works, your own personal perception of it, that leads you to say that. If all you had was the mathematics of quantum mechanics, but no conscious experience to draw on and no self-modeling in your analysis, you would never have any reason to think that your statement above was true. It just plain does not come from the mathematics of quantum mechanics, so where does that expectation come from?

What? But everytime there is a hit detected in the screen in the double slit. It is definite outcome where one of the eigenvalue of positions is chosen. The robot with even a $2 vga camera vision can see the hit. What's the problem with this??

 

[Ken G]

What? But everytime there is a hit detected in the screen in the double slit. It is definite outcome where one of the eigenvalue of positions is chosen. The robot with even a $2 vga camera vision can see the hit. What's the problem with this??

The problem is that you have no evidence at all that it is true, independently of your conscious experience. There is nothing in the theory of quantum mechanics that suggests this would be true. So it's something you are intentionally tacking onto the theory, and why are you doing that? Because your personal conscious experience tells you to do that, no other reason. Certainly nothing that robots would have access to, they could be programmed to function just fine in the language of mixed states, and they could test the mixed states the same way all quantum mechanical predictions are tested: in the classical limit. Even as the robots are functioning, you are imagining "checking in on them", using your conscious experience-- either in practice or hypothetically to form the words you are using. You have to remove all your conscious experience from the claims you are making, if you are to claim they hold in the absence of conscious experience. This is of course very hard to do, so we don't do it-- but we can still recognize that we are not doing it.

 

[Varon]

The problem is that you have no evidence at all that it is true, independently of your conscious experience. There is nothing in the theory of quantum mechanics that suggests this would be true. So it's something you are intentionally tacking onto the theory, and why are you doing that? Because your personal conscious experience tells you to do that, no other reason. Certainly nothing that robots would have access to, they could be programmed to function just fine in the language of mixed states, and they could test the mixed states the same way all quantum mechanical predictions are tested: in the classical limit. Even as the robots are functioning, you are imagining "checking in on them", using your conscious experience-- either in practice or hypothetically to form the words you are using. You have to remove all your conscious experience from the claims you are making, if you are to claim they hold in the absence of conscious experience. This is of course very hard to do, so we don't do it-- but we can still recognize that we are not doing it.

Is the above argument the position of other Copenhagenists too, or is it your particular only? It's definitely a weird argument. I'll ponder on it.

 

[BruceW]

Ken G - For the case of the mindless robot in the world without any consciousness:
Us humans are classical objects that can cause a non unitary process. Therefore, I would deduce that other similar classical objects which do not have consciousness may also cause a non unitary process. I would not, like you, deduce that non-conscious beings cannot cause a non unitary process.
Therefore, in the world with no conscious beings, I would say that non unitary processes still happen.

 

[BruceW]

(That's for the Copenhagen interpretation). And for many-worlds theory, neither conscious nor non-conscious objects cause a non unitary process.
So neither of these theories make any distinction between conscious and non-conscious objects.

 

[Varon]

Ken G - For the case of the mindless robot in the world without any consciousness:
Us humans are classical objects that can cause a non unitary process. Therefore, I would deduce that other similar classical objects which do not have consciousness may also cause a non unitary process. I would not, like you, deduce that non-conscious beings cannot cause a non unitary process.
Therefore, in the world with no conscious beings, I would say that non unitary processes still happen.

If I understand it right. I think what Ken G was saying was that humans have self initiated volition to home in on the definite outcomes that machines or even equations in quantum mechanics don't have. It's like the following is the case:

LEVELS -------- CAPABILITY
microworld = superposition
lifeless robots = mixed state
humans = definite outcome

Now how come or what collapse the wave function. If Copenhagen was right. Wave function could be collapsed to definite outcome because of the existence of self-initiated volation. Without this possibility. Nature would only choose up to mixed state and no possibility of definite outcome. So it's like some kind of anthropic principle why nature has this capability to collapse wave function.

I guess this is all there is to it to Copenhagen, a temporary system of philosophy for lack of other data that can lock on the complete theory and correct interpretation.

 

[BruceW]

Going back to Fredrik's last post, he is right that the definition of a classical measuring apparatus is simply that it cannot practically be 'reversed' to the state it was in before the measurement.
I realize this is a rather vague definition. But I would rather say that all classical measuring devices cause a non unitary process, rather than say that conscious beings cause a non unitary process.

 

[Fredrik]

Varon (and BruceW), Ken G isn't saying that a human observing something is any different from a robot doing it. He's saying that the concept of measurement involves consciousness, not that the act of measurement does.

A human would never think of an interaction that leaves no record of the result as a "measurement". The reason is that he wouldn't have any way of knowing what the result was. If he knows that there's been a result, then a record has been created, because his knowledge is a record.

 

[Varon]

Varon (and BruceW), Ken G isn't saying that a human observing something is any different from a robot doing it. He's saying that the concept of measurement involves consciousness, not that the act of measurement does.

A human would never think of an interaction that leaves no record of the result as a "measurement". The reason is that he wouldn't have any way of knowing what the result was. If he knows that there's been a result, then a record has been created, because his knowledge is a record.

So a robot can be considered the same level as human doing "measurement" if the robot has ability to register a single outcome to an experiment... which is what is attributed to the word "consciousness". I think we must not use the word "consciousness" to avoid unnecessary troubles and miscommunications. Just use other terms.

 

[nickelite]

You are also talking about the easy problem of collapse, the decoherence. That does not tell us how we get a single outcome, that question remains entirely unanswered and very much does require an interpretation choice, or you have to punt the question entirely (which is more or less choosing CI interpretation).

But either way, there is no experiment in the near future that will be able to distinguish between our two interpretations.

collapse, in my opinion (its all opinions here anyway

I invite you to watch this video. Since I have no way of doing the experiment and no one I know has, my question is, Do you think that what they say in it is true?

Wouldn't that experiment be conclusive on whether consciousness is involved? And really differentiate opinions from facts? I mean whether the results of the experiments in the video are true or not, I still have my doubts about it.Link: http://www.youtube.com/watch?v=_OWQildwjKQ&feature=related"

Also in the Feynman Lectures Vol 3. They imply that the cause of collapse is interaction of the electrons with the light source.

Do you think it's the photons? Isn't there a way to determine if it's them, or maybe another factor?

Cheers!

 

[Ken G]

Ken G - For the case of the mindless robot in the world without any consciousness:
Us humans are classical objects that can cause a non unitary process. Therefore, I would deduce that other similar classical objects which do not have consciousness may also cause a non unitary process. I would not, like you, deduce that non-conscious beings cannot cause a non unitary process.

We have two choices. We can notice that we are conscious, and we are classical, and just guess which of those attributes causes us to experience non-unitary phenomena. Or, we can say, if we were classical but non-conscious, would we need the non-unitary explanations? I've argued we would not. If we were conscious but non-classical, would we need the non-unitary explanations? I would say yes we would, because if one can imagine a conscious electron, it still has to register a single outcome to an experiment. That is the fundamental nature of consciousness, its association with experience. I probably could have ducked a lot of this difficulty by just replacing "consciousness" with "experiential agent", because I'm using the terms synonymously.

Therefore, in the world with no conscious beings, I would say that non unitary processes still happen.

And I claim that is because even in that world with no conscious beings, you are implicitly slipping in hypothetical conscious agents, to help you form your language about that world.

 

[Ken G]

(That's for the Copenhagen interpretation). And for many-worlds theory, neither conscious nor non-conscious objects cause a non unitary process.
So neither of these theories make any distinction between conscious and non-conscious objects.

But in many worlds, we still have to place conscious experience, and it is the reason we talk about "many" worlds instead of just one unitary world. So the many-worlds interpretation is the one that has the most apparent of all roles of consciousness as the source of non-unitary experience, it's just crystal clear if one adopts that interpretation.

 

[Ken G]

If I understand it right. I think what Ken G was saying was that humans have self initiated volition to home in on the definite outcomes that machines or even equations in quantum mechanics don't have.

Yes, that's it exactly.

LEVELS -------- CAPABILITY
microworld = superposition
lifeless robots = mixed state
humans = definite outcome

Right. Indeed, I'm essentially defining consciousness as "the experience of a single outcome", because that's the only kind of consciousness we know about. The way we experience our own consciousness is in a kind of necessity to have a single outcome. So we have to shoehorn the outcome of any experiment into that rubric. I don't know if we do that because we are conscious, or if we are conscious because we can do that. Thinking about what consciousness actually is gets quite difficult, I am merely looking at what ramifications having it presents us with.

Now how come or what collapse the wave function. If Copenhagen was right. Wave function could be collapsed to definite outcome because of the existence of self-initiated volation.

That's not how the CI frames it. In the CI, the wavefunction is just a calculational tool, so its final "collapse" to a definite outcome is an entirely manual process. It is just a thought, and you don't have to "collapse" thoughts, you just think them. Ironically, CI manipulates the wavefunction in exactly the way you like to imagine the mind can manipulate reality, making it a closer cousin to your way of thinking than you realize.

Without this possibility. Nature would only choose up to mixed state and no possibility of definite outcome. So it's like some kind of anthropic principle why nature has this capability to collapse wave function.

Yes, except I don't see anthropic principles as principles of nature, I see them as ramifications of human intellect. Nature imposes constraints on the intellect by virtue of what thought processes are successful, but the thought processes themselves are from us, not from nature outside of us.

I guess this is all there is to it to Copenhagen, a temporary system of philosophy for lack of other data that can lock on the complete theory and correct interpretation.

CI would say there is no complete theory nor correct interpretation in the sense you mean, so it asserts itself as the next best thing. It simply avoids self-delusion, that's its raison d'etre.

 

[Ken G]

A human would never think of an interaction that leaves no record of the result as a "measurement". The reason is that he wouldn't have any way of knowing what the result was. If he knows that there's been a result, then a record has been created, because his knowledge is a record.

Spot on. There is something inseparable between the act of being conscious, and the ability to have a unique record of a happening, but I don't know which is prior to the other. Maybe the two are one in the same in some sense.

 

[Varon]

We have two choices. We can notice that we are conscious, and we are classical, and just guess which of those attributes causes us to experience non-unitary phenomena. Or, we can say, if we were classical but non-conscious, would we need the non-unitary explanations? I've argued we would not. If we were conscious but non-classical, would we need the non-unitary explanations? I would say yes we would, because if one can imagine a conscious electron, it still has to register a single outcome to an experiment. That is the fundamental nature of consciousness, its association with experience. I probably could have ducked a lot of this difficulty by just replacing "consciousness" with "experiential agent", because I'm using the terms synonymously.

That's right. If you just use "experiential agent" instead of "consciousness", you could
have ducked a lot of this difficulty. "Consciousness" is a fully loaded word depending on the awareness of the person talking about it. It has dozens of definitions. Let's go to your "experiential agent", should the "experiential agent" be self-aware? What is the minimum requirement for it?

And I claim that is because even in that world with no conscious beings, you are implicitly slipping in hypothetical conscious agents, to help you form your language about that world.

 

[Ken G]

Let's go to your "experiential agent", should the "experiential agent" be self-aware? What is the minimum requirement for it?

The minimum requirement is the capability and even necessity to conceive of a single outcome for an experiment. We have to register a unique experience. In classical physics, that seems like a perfectly normal thing for any material object to do, and even though in statistical mechanics the fundamental object is a probability distribution, we still imagine we have an ensemble of single outcomes. But in quantum systems, the formal mathematics of quantum mechanics makes no mention of single outcomes to experiments, and single outcomes are never predicted and never tested. So this is an important "disconnect" for conscious entities. I'm arguing it is more important than the quantum/classical gap that Bohr and co. tended to focus on, because the two are so easily mistaken for each other. Indeed, if you think about it, the classical limit is always the thermodynamic limit, so we really always have probability distributions rather than single outcomes, even in classical physics-- it's usually just lost in the measurement error.

 

[Varon]

The minimum requirement is the capability and even necessity to conceive of a single outcome for an experiment. We have to register a unique experience. In classical physics, that seems like a perfectly normal thing for any material object to do, and even though in statistical mechanics the fundamental object is a probability distribution, we still imagine we have an ensemble of single outcomes. But in quantum systems, the formal mathematics of quantum mechanics makes no mention of single outcomes to experiments, and single outcomes are never predicted and never tested. So this is an important "disconnect" for conscious entities. I'm arguing it is more important than the quantum/classical gap that Bohr and co. tended to focus on, because the two are so easily mistaken for each other. Indeed, if you think about it, the classical limit is always the thermodynamic limit, so we really always have probability distributions rather than single outcomes, even in classical physics-- it's usually just lost in the measurement error.

Can we program a lifeless robot to conceive of a single outcome for an experiment? Or must the robot be self aware?

 

[Ken G]

Can we program a lifeless robot to conceive of a single outcome for an experiment? Or must the robot be self aware?

I'm not sure one can answer that, one can only speak of what they mean by the words. If we did program a robot to be able to conceive of and experience a single outcome, then for the purposes of everything I've said above, that robot would be conscious. Whether or not it is possible to actually program a robot to be conscious is a question that is debated ad nauseum in the artificial intelligence community, and no one strikes me as being terribly close to a definitive answer, nor does it matter at all to anything I've said because the "robots" I talked about were ones that we could actually imagine programming with existing technology.

 

[Varon]

I'm not sure one can answer that, one can only speak of what they mean by the words. If we did program a robot to be able to conceive of and experience a single outcome, then for the purposes of everything I've said above, that robot would be conscious. Whether or not it is possible to actually program a robot to be conscious is a question that is debated ad nauseum in the artificial intelligence community, and no one strikes me as being terribly close to a definitive answer, nor does it matter at all to anything I've said because the "robots" I talked about were ones that we could actually imagine programming with existing technology.

You said the lifeless robot can only perceive mixed state in the double slit experiment. Can you please be clear what is the equivalent of mixed state in the double slit. I think it has to do with the fact that we don't know which slit the photon would enter. But won't this produce detection events? I guess not. So what does it mean the robot can only perceive mixed state when nothing in the detector got any photon hits?

 

[Ken G]

You said the lifeless robot can only perceive mixed state in the double slit experiment. Can you please be clear what is the equivalent of mixed state in the double slit. I think it has to do with the fact that we don't know which slit the photon would enter. But won't this produce detection events? I guess not. So what does it mean the robot can only perceive mixed state when nothing in the detector got any photon hits?

The robot doesn't perceive at all, that's the point. What we are talking about is a program and a machine running that program, and its job is to test quantum mechanical predictions. Quantum mechanical predictions are mixed states, that's it, that's all you get from quantum mechanics. So the robot must be set up to test the mixed-state predictions, and this is always done in the classical limit. That's the ultimate irony of doing quantum science, all the predictions of the quantum theory are tested in the classical limit of many events, but they can be done one at a time and tested with outcomes that are quantum in nature. However, the outcomes that come from quantum mechanics are mixed states, so a robot set up to test quantum mechanics in a universe with no perception and no concept of a unique outcome would have to test quantum mixed states. The tests come out the same as classical ensembles, so rarely are the quanta sent through one at a time. Instead, the robot would be programmed to recognize the superposition principle, and would send a vast ensemble through their instrument, and test the mixed-state predictions in the classical limit, and find that quantum mechanics works with flying colors. All without ever having a concept of a single unique outcome for a single quantum.

Now, a robot might also be set up to test if the classical limit also holds if the quanta are sent through one at a time. Again quantum mechanical predictions look like probabilities of certain quantized events occurring when you project onto the quantum, but when you think in terms of the whole system making the measurement, it is still a pure state. You can't test the pure state of the whole system, it's too complicated, so you have to test the mixed state of the quantum subsystem, and that's where the probability distribution comes in. But there's still no concept of a single outcome here, if we believe the prediction of the theory, there is a probability distribution of outcomes. That's a very bizarre ontological entity, our minds can only handle it by replacing it with an ensemble of single outcomes that divide up according to the probability fractions (the old "one thing happened by I don't know which" construct). But the predicted outcome of the theory, with no intelligences there, is not actually that, instead it is a superposition of the instrument registering single outcomes, but no single global outcome in the reality, those single outcomes are not actualized individually, they are part of a superposition.

At first when you just have a single run of the apparatus, the terms in the superposition look very different, but when you run it over and over and aggregate the results, they all start to look the same. A superposition of things that are all the same is just one thing, and so in the classical limit, the "single reality" is recovered, and that single reality is a robot attesting to the success of the prediction. Never is there a need for any artificial imposition of a single global reality, until one gets to the classical limit where the theory can really be tested accurately. Unless there are consciousnesses at those intermediate stages.

 

[Varon]

That's simply because the phrase "consciousness causes collapse" can mean about 100 different things, many of which have been carefully delineated on this thread. Indeed that was the initial disconnect between myself and G01, if you care to review that discussion. Nothing that I am saying is what that Wiki means by consciousness causing collapse, but what I do mean is completely consistent with both the Wiki I quoted for you, and what G01 just said above.

All covered in the thread, you're not telling me anything I haven't already discussed in detail.
Now I'm just repeating, but I will go through it yet again because this is the crux of the whole business right here.

A measurement made by a lifeless computer creates what is called a mixed state for the quantum system. That is because the quantum system is, at this point, a subspace of a larger apparatus. The larger apparatus, as G01 just explained, is still in a pure state according to the unitary evolution of quantum mechanics theory. The mixed state is the projection from the whole system onto the subspace of the quantum system. Nothing there creates any difficulties, nor requires any interpretations of quantum mechanics-- the full system is in a pure state so is still unitary, the projection onto the subspace is not supposed to be unitary, it's a projection from a joint wave function to a single-particle state, and that does not lead to a single-particle wavefunction at all (let alone an eigenstate of the measurable), it leads to a mixed state.

After your last message, it was clearer that the "lifeless robot" can only see the mixed state after numerous runs of the single photon emission forming an ensemble. Realizing this. I went back and forth in this thread over an hour reading old messages to get the whole context. So let me go back to this. When exactly is a double slit in pure state? I understand well that when a quantum system is completely isolated, it is in pure state, and anytime a subspace or region is focus due to decoherence, it is in mixed state. How do you apply this to the double slit setup? This is because you are talking of ensemble of many million single photon emission. Where is the pure state here? When only one photon is sent a time? But you said mixed state is equal to numerous single photon experiments, so is pure state related only to one photon at a time or can you consider pure state in some scenerio of the numerous runs?

Now back to the "lifeless robot", you said it can't perceive single or definite outcome, but note that each hit in the detector is composed of single outcome. Even though the lifeless robot can only detect the ensemble, but an ensemble is made up of each one photon at a time emission. So you can't say that a "lifeless robot" or as you put it below "measuring devices should not "cause a non unitary process to occur." Remember each photon at a time emission and detection is already a non unitary process. The lifeless robot or measuring device can detect ensemble of them.. in other words, it can detect ensemble of millions of single non-unitary process... although only human consciousness can perceive the intermediate and know it is because of the single non-unitary process. Pls. be precise in your language, then BruceW may completely understand you too because I feel he gives up understanding it because of the critical issues I mentioned above. Thanks.

At this point, where all we have is the "lifeless computer", we do not have a single measurement outcome, we have a mixture of outcomes. This is also called an "ensemble" in mainstream quantum mechanics, the only difference is that to resolve certain difficulties in picturing what this is, we imagine lots and lots of copies of the system, instead of just one system. This makes it easier to imagine what a mixed state is, but there's really only one system there, it doesn't have to be an ensemble.

Enter a consciousness/intelligence/perceiver who thinks classically. Only now do we encounter the concept of an "actual outcome", and this creates a huge problem for quantum mechanics theory. Where does that actual outcome come from? No one knows, but here is where each of the interpretations step into provide an untestable answer. I've already outlined what those answers are above, and GO1 mentioned some of the possibilities as well. The key thing to recognize at this point is that none of that difficult business even comes up, and there's no need for an interpretation, until we factor in the presence of a consciousness and its resulting "actual outcome" perception. The physics is perfectly happy just leaving the quantum in a mixed state, if all we have is a lifeless computer. It's all related to how a conscious entity does science, and this involves the perception of an actual outcome, even though the theory provides no such concept and forces us to inject a layer of randomness to get agreement with our experiences. Because we are conscious.

So this role of consciousness is much more subtle, yet much more fundamental to everything we do in science, than what that Wiki is talking about. I know that without even reading the context of the rest of that Wiki.

No. Measuring apparatuses are physical constructs, and so, should obey the laws of physics. Unitary evolution is one of the laws they should obey. Ergo, measuring devices should not "cause a non unitary process to occur." The whole measurement problem, as nicely described by G01 above, is the origin of the apparent non-unitarity, since it cannot come from the measuring device. Here are the answers of the main interpretations:

CI: it comes from how we do science, since the unitary evolution piece was just a tool we use at one stage of the calculation. (This is related to our consciousness/intelligence/classical processing in the "how we do science" part.)

Many-worlds: the non-unitary element is illusory, the full unitary result is there but we only see a tiny part of the real story. (This is related to our consciousness/intelligence/classical processing in the "what part of the whole we see".)

Bohm: the unitary evolution is the illusion-- it's just a cloak placed on top of the pilot-wave evolution, which is deterministic and non-unitary. The unitarity is "filled in" by physically irrelevant aspects of the wave function, and it is stripped away by the measurement. (This is the only interpretation that does not involve consciousness in a direct way, because it treats unique experimental outcomes as purely deterministic, but it cannot be tested.)

 

[BruceW]

If we were conscious but non-classical, would we need the non-unitary explanations? I would say yes we would, because if one can imagine a conscious electron, it still has to register a single outcome to an experiment. That is the fundamental nature of consciousness, its association with experience. I probably could have ducked a lot of this difficulty by just replacing "consciousness" with "experiential agent", because I'm using the terms synonymously.

No, an electron in quantum state superposition will not undergo a non-unitary process. Only classical objects can undergo a non-unitary process.
(This is why decoherence is important. Non-unitary process can only happen after decoherence, but a single electron cannot 'decohere'. Only a large system of many particles can decohere).

But anyway, back to the main subject,
The postulates of QM say that there is a superposition of quantum states, then measurement causes an outcome of one of those states to happen with a particular probability. Then after measurement, the wavefunction will be equal to the state corresponding to the outcome.
(This is a measurement of the first kind, aka strong von Neumann projection, aka full collapse, ie the non-unitary process has happened).
From this method, you don't have to set up the system so that a person is making the measurement. Any classical object can 'make a measurement'.
It is true that if we want to predict the outcome for a human, the thing making the measurement can always be chosen to be human (since this is always the last step in any experiment). But generally, we can choose the non-unitary process to happen at the lab equipment, instead of at the person.

I am going with the general case that any classical object can cause the non-unitary process.
Ken G's opinion (as far as I understand) is that since we only need to find predictions for what happens from a human's perspective, we can choose our definition of 'measurement' such that only a human can make a measurement.

Ultimately, you could go one step further and say that the only perspective I am interested in is my perspective. I can then define 'measurement' such that only I can make a measurement. Therefore, when my scientist friend does an experiment, the outcome of the experiment (and my friend) are all just a quantum superposition, until he comes to see me, and then by seeing him, I cause a measurement to happen. This interpretation would work equally well.

So when Ken G was saying 'conscious', in my words, I would say 'the perspectives which we are interested in finding predictions for'. Or more precisely, 'classical objects which we define to be able to make measurements'.

I had to think for a long time to get this far. You're right, Varon, I had great difficulty understanding what conscious meant in this context.

 

[Varon]

No, an electron in quantum state superposition will not undergo a non-unitary process. Only classical objects can undergo a non-unitary process.
(This is why decoherence is important. Non-unitary process can only happen after decoherence, but a single electron cannot 'decohere'. Only a large system of many particles can decohere).

But anyway, back to the main subject,
The postulates of QM say that there is a superposition of quantum states, then measurement causes an outcome of one of those states to happen with a particular probability. Then after measurement, the wavefunction will be equal to the state corresponding to the outcome.
(This is a measurement of the first kind, aka strong von Neumann projection, aka full collapse, ie the non-unitary process has happened).
From this method, you don't have to set up the system so that a person is making the measurement. Any classical object can 'make a measurement'.
It is true that if we want to predict the outcome for a human, the thing making the measurement can always be chosen to be human (since this is always the last step in any experiment). But generally, we can choose the non-unitary process to happen at the lab equipment, instead of at the person.

I am going with the general case that any classical object can cause the non-unitary process.
Ken G's opinion (as far as I understand) is that since we only need to find predictions for what happens from a human's perspective, we can choose our definition of 'measurement' such that only a human can make a measurement.

Ultimately, you could go one step further and say that the only perspective I am interested in is my perspective. I can then define 'measurement' such that only I can make a measurement. Therefore, when my scientist friend does an experiment, the outcome of the experiment (and my friend) are all just a quantum superposition, until he comes to see me, and then by seeing him, I cause a measurement to happen. This interpretation would work equally well.

So when Ken G was saying 'conscious', in my words, I would say 'the perspectives which we are interested in finding predictions for'. Or more precisely, 'classical objects which we define to be able to make measurements'.

I had to think for a long time to get this far. You're right, Varon, I had great difficulty understanding what conscious meant in this context.

I think what Ken was saying was simply that both machines and human consciousness are similar in that wave function collapse behavior was similar in their presence. What differs is how we perceive it. Since the only way to create an ensemble to match the prediction of the equation is to run many one at a time photon emission. The machine can only see the interference pattern after millions of single photon emissions as ensemble. Intermediate. It can't perceive the pattern as cause by collapse because machines can only see ensemble because the equation output can only be demonstrated with ensemble. It is only human which can know that a single outcome or hit in the detector is because of non-unitary process in a single photon at a time run. I think this is what Ken was saying. If I'm wrong. Let me know.