April 7th, 2010

I just gave a talk at the Philosophical Seminar workshop (hosted by Gilles Kuhn) of the Kira Institute (started by various academics associated with Institute for Advanced Study, Stanford, Amherst, Princeton, and others), in Second Life on the subject of an approach to an interpretation of quantum mechanics which my friend Jonathan Tash and I came up with. My name is “Mitsu Ishii” in Second Life:

Gilles Kuhn: hello melchi
Mitsu Ishii: hello
Kendust Vansant: hello everybody
Gilles Kuhn: hello kendust take a seat
Melchizedek Blauvelt: Hi Gilles, all
Bleu Oleander: hi
Mitsu Ishii: all right, so shall I begin?
Gilles Kuhn: ok so in this seminar we will “hear” a presentation of mitsu and yes please by all means do so
Kendust Vansant: my first visit here
Mitsu Ishii: Okay. So I assume most of you have spent some time already studying various interpretations of quantum mechanics
Mitsu Ishii: As you know, there is a fundamental problem, the measurement problem, which lies at the heart of QM.
Agatha Macbeth: Hi Gilles
Mitsu Ishii: hello Agatha
Gilles Kuhn: helo agatha
Bleu Oleander: hi Agatha
Agatha Macbeth: Hello Bleu
Mitsu Ishii: I’ll start with a brief reprise of two archetypical interpretations which I think can serve as grounding points.
Mitsu Ishii: So, first the Copenhagen Interpretation states that quantum wavefunctions which evolve in a unitary fashion “collapse” when subjected to observation.
Mitsu Ishii: So, for example, you have a wave function which exists as a superposition of, say, a particle spinning clockwise and the same particle spinning counterclockwise (spin up versus spin down).
Mitsu Ishii: We can only actually observe the particle in one or the other orientation of spin; it is either spin up or spin down, in other words.
Mitsu Ishii: Copenhagen says, the particle is in a superposition of both states until it is observed, at which point it “collapses” into one or the other state, entirely randomly.
Mitsu Ishii: The fundamental difficulty with this interpretation is brought out by Schrodinger’s Cat which I’m sure you’re all familiar with.
Mitsu Ishii: Briefly, to an external observer, the whole system: observer, particle, etc., is itself in a superposition. There’s “observer observes spin up” and “observer observes spin down” and until the external observer “looks into the lab” the lab, including observer and particle, is in a superposition of both states.
Mitsu Ishii: so the problem with Copenhagen is that there seems to be an infinite regress — who or what constitutes an “observer” ?
Mitsu Ishii: One possible objection is that in Nature it appears that only microscopic objects can be in superposition, though the theory says that there’s nothing that stops macroscopic objects from existing in superposition.
Mitsu Ishii: However, most recently there has been an experiment showing a macroscopic object, visible to the naked eye, about the width of a human hair, existing in superposition.
Mitsu Ishii: There’s no technical reason why that could not be expanded to larger and larger objects.
Mitsu Ishii: So at this point I’d like to mention the Everett Interpretation, which is one that many cosmologists favor, according to polls.
Mitsu Ishii: In the Everett Interpretation, there is no objective wavefunction collapse at all.
Mitsu Ishii: This interpretation is also called the “relative state interpretation”.
Mitsu Ishii: The basic idea in Everett is that rather than having an objective collapse, there is simply a correlation between a mind state and an observation.
Mitsu Ishii: That is to say, the particle itself remains in a superposition of spin up and spin down, but there is a mental state which is correlated with spin up and another mental state correlated with spin down.
Mitsu Ishii: Colloquially, it’s said that the universe “splits into two” universes, one with a measurement of spin up and one with a measurement of spin down
Mitsu Ishii: however this is somewhat misleading. What is more accurate is to say that the measurement of spin up is correlated with one mind, and spin down with another mind, but measurements are all relative to a specific observer/mind.
Mitsu Ishii: One could say, for example, that in this interpretation, when we look at the past, the past is not objectively present before a mind correlates with it.
Mitsu Ishii: so the dinosaurs, etc., don’t exist prior to digging up the fossil evidence, which then correlates your mind with the retrojectively projected past which has dinosaurs, and so on.

Mitsu Ishii: This is a point Lee Smolin has made, for example, in his popular book on quantum gravity.
Mitsu Ishii: Now, there is still a fundamental missing component to the Everett Interpretation.
Mitsu Ishii: One problem is, Everett does not define what a “mind” is, or how the mind correlates with an observation.
Mitsu Ishii: It’s simply assumed that some sort of correlation occurs, somehow, with a mind.
Mitsu Ishii: The second, and really fundamental problem, is called the preferred basis problem.
Mitsu Ishii: I’m not sure how familiar you are with the mathematics, so I’ll just introduce this in as simple a way as I can.
Mitsu Ishii: If you think about referencing a point on a plane, you can imagine using Cartesian coordinates.
Mitsu Ishii: so this point is 2 units along the x axis and 1 unit along the y axis.
Mitsu Ishii: however, obviously one can choose any non-parallel pairs of vectors in the plane to set up your coordinate system.
Mitsu Ishii: for example, you can “rotate” the basis vectors by 45 degrees. That will change the coordinate representation of all your points in a straightforward way.
Mitsu Ishii: In quantum mechanics, you can represent the state of a system in a similar fashion.
Mitsu Ishii: For example, a particle in a spin up and spin down superposition can be represented by a vector: spin up + spin down (times a scaling factor for normalization)
Mitsu Ishii: you can think of “spin up” as like the x-axis, and “spin down” like the y axis.
Mitsu Ishii: when the particle is measured, however, it seems to “collapse” and change state to either a pure spin up or pure spin down vector in state space.
Mitsu Ishii: Now the big mystery is (drum roll…)
Mitsu Ishii: why is it that when we observe things, we only observe them collapsed into a particular “preferred” basis?
Birric Forcella: I have a solution for that but you are not going to buy it
Mitsu Ishii: In other words, if you accept the Everett Interpretation, why is it that minds only correlate with “spin up” or “spin down” but never “spin up + spin down” or any other vector?
Mitsu Ishii: Okay, we can discuss your solution in a minute, Birric But first I’ll lay out my proposed solution…
Birric Forcella: No, next time or so. You go on
Mitsu Ishii: Okay.
Mitsu Ishii: I hasten to add that I have not worked out this solution — it is more a research direction that my friend Jonathan Tash and I came up with a while ago.
Mitsu Ishii: We spent a week at a Kira Summer School working on it but it remains a work in progress.
Mitsu Ishii: But I think it is fairly interesting so I’ll present what we have, now.
Mitsu Ishii: One of the interesting facts about the preferred basis is that it has been proven that what we call “locality” can only exist in a universe where observations are always made relative to the preferred basis.
Mitsu Ishii: So, although mathematically there seems to be no reason why minds correlate only with observations made in pure, unmixed states relative to the preferred basis,
Mitsu Ishii: a universe in which one could observe mixed states directly would lose any notion of local spacetime.
Mitsu Ishii: As you may know, mixed quantum states can involve particles that are separated across the universe, and so on.
Mitsu Ishii: quantum entangled particles, as far as we know, remain entangled even when separated by thousands of light years.
Mitsu Ishii: If we could directly observe mixed states, ordinary notions of locality in spacetime would be violated.
Mitsu Ishii: So, it occurred to Jonathan and I that one approach to solving this problem would be to reassess the definition of “mind” — something which is more or less left out of nearly all attempts to address the quantum interpretation problem.
Gilles Kuhn: by ordinary notion of locality i presume you mean lorentz invariant?
Mitsu Ishii: What I mean, essentially, is that, for example, when we observe a cup, it is a cup that is local, i.e., it’s right here, in front of us.
Mitsu Ishii: We do not have to think about the cup here in superposition with a cup on the other side of Jupiter.
Mitsu Ishii: or on the other side of the universe.
Mitsu Ishii: or superimposed with cups scattered all across the universe, and so on.
Mitsu Ishii: the preferred basis allows us to “observe” things that appear to be right here, in one location.
Mitsu Ishii: we can of course indirectly infer superpositon via many means, but it appears we cannot directly observe objects or particles in mixed states.
Mitsu Ishii: so the question is, why is this?
Mitsu Ishii: so anyway, the idea here is, what is a mind? and Jonathan suggested we might think of this in evolutionary terms.
Mitsu Ishii: Perhaps there is an evolutionary advantage to observing things in the preferred basis.
Mitsu Ishii: Then it occurred to me to think about the work of the cyberneticist Gregory Bateson.
Mitsu Ishii: I won’t go into his work in great detail but I highly recommend the book _Mind and Nature_. It is really a revelation in many ways.
Mitsu Ishii: But basically, Bateson was trying to understand life systems and systems with “mental” qualities entirely within the paradigm of classical physics.
Mitsu Ishii: He ended up coming up with a definition that is quite simple.
Mitsu Ishii: By thinking carefully about mental systems, he realized that they all involve feedback loops of some kind.
Mitsu Ishii: By “mental” he means basically living systems, not just “minds” in the sense we usually imagine.
Mitsu Ishii: For example, every homeostatic system involves a feedback loop. He used a heater-thermostat as a very simple example of a homeostatic system (he thinks artificial and biological living systems are equivalent from his point of view, if they share certain characteristics).
Mitsu Ishii: it gets warm, the thermostat switches off, it cools down, the thermostat switches on, and so forth, in a loop.
Mitsu Ishii: There is a feedback between the system and the environment.
Agatha Macbeth nods
Mitsu Ishii: the thermostat is a sensor which interacts with the heater which interacts with the environment which interacts with the thermostat.
Mitsu Ishii: in his view, you need to have three or more elements interacting in a circular or more complex feedback loop to get what he calls “phenomena of mind”
Mitsu Ishii: There are lots of interesting things he points out in the book relative to this question.
Mitsu Ishii: One thing he points out is that because of the feedback, mental systems are self-referential.
Mitsu Ishii: the circularity of the feedback creates a kind of information self-reference.
Mitsu Ishii: he pointed out that many logical paradoxes occur also as a result of feedback loops.
Mitsu Ishii: Russell’s paradox, Godel’s Theorem, etc., all involve self reference.
Mitsu Ishii: He asserted that one reason we have difficulty understanding phenomena of mind is that our logical system is designed to map onto an external world that we conceptualize without feedback loops.
Kendust Vansant: thank you mitsu, bye all!
Mitsu Ishii: by Kendust
Mitsu Ishii: bye
Mitsu Ishii: So for example, with a thermostat, you could write “it is hot, therefore the thermostat is off. If the thermostat is off, it gets cool. if it is cool, the thermostat is on.”
Mitsu Ishii: if you don’t include a time delay in this, it looks like a self-contradictory set of assertions.
Mitsu Ishii: “The thermostat is off, therefore the thermostat is on.”
Mitsu Ishii: He made the interesting point that when physicists were first trying to write down thermodynamic equations for the steam engine, they could solve the equations.
Mitsu Ishii: The steam engine has three parts that interact in a circular feedback loop.
Mitsu Ishii: But the thermodynamic equations between the parts of a steam engine result in a mutually inconsistent set of equations.
Mitsu Ishii: Maxwell was the first physicist to understand that the reason for this is that the equations are equilibrium equations.
Mitsu Ishii: but in a real steam engine there is a slight time delay from a change in one part of the engine to the next.
Mitsu Ishii: So in fact you can solve the equations, but only if you include the time delay.
Mitsu Ishii: Okay, so back to quantum mechanics (I am nearly done!)
Mitsu Ishii: So the basic idea of our interpretation is that what is going on with “observation” is a circular feedback loop between elements.
Mitsu Ishii: In order for this feedback loop to “close”, however, it has to be comprised of feedback loops which can complete in finite time.
Mitsu Ishii: Thus, if we were able to observe mixed states, which could conceivably involve superpositions of states which are radically nonlocal
Mitsu Ishii: there would be no guarantee of being able to have the feedback loops close in a finite time, which is required for computation (mental phenomena).
Mitsu Ishii: Thus, we hypothesize that the preferred basis is selected out by the architecture of mental phenomena which require feedback loops a la Gregory Bateson.
Mitsu Ishii: A corollary to this is that what we call “space” and “time” are artifacts of perception.
Mitsu Ishii: Time is literally created, in our view, by the operation of mind (this is a somewhat Kantian idea as well).
Mitsu Ishii: space as well.
Mitsu Ishii: So in fact objects like rocks and planets and even stars exist, in this view, relative to living systems which require them to backfill a universe that allows perception in their local quantum context.
Mitsu Ishii: i.e., just as dinosaurs are ‘created’ when you dig up the fossils, the physical universe is “created” as a backdrop, so to speak, for local observational loops.
Mitsu Ishii: this is not a solipsistic interpretation, however, because there can be cross-cutting intersubjective agreement between different feedback loops.
Mitsu Ishii: so this creates what appears to be a shared objective world to some degree.
Mitsu Ishii: But the fact that it is not entirely objective is what creates what we observe as quantum phenomena, superposition, and so on.
Mitsu Ishii: So to conclude: the idea is that the classical world is in fact the “strange” world — the quantum world is the “normal” world. it is living systems (including any scale of living system) which in some sense induce a quasi-classical apparent world.
Mitsu Ishii: That’s our basic idea.
Mitsu Ishii: Sorry it took so long, I’d be happy to answer questions or discuss now or at a later time.
Katharine Kozlowski is Online
Gilles Kuhn: clap clap clap thank you for this very interesting presentationb mitsu
Agatha Macbeth: Domo arigato Mitsu
Gilles Kuhn: well floor is open to question observation and criticism
Birric Forcella: clap clap clap
Birric Forcella: Really nice
Bleu Oleander: phew!
Wester Kiranov: ty mitsu, very interesting - have a book to read I think
Zen Arado: tnaks Mitsu
Bleu Oleander: ty Mitsu!
Mitsu Ishii: thanks. yes, feel free to ask me questions now or later, or offer critique. As I say it’s just the beginning of a research direction.
Mitsu Ishii: I wish I had more time to work on it
Agatha Macbeth: Sounds like a good beginning
Wester Kiranov: this ‘observation’ idea you have now is a lot broader than ususal ideas in that direction
Wester Kiranov: basically anything witha complex enough feedback loop, right?
Gilles Kuhn: well i have a general question about the “nature” and use of mind but to pinpoint it i want to stress that if i remember well for bateson any kind of cybernetic looping system is a “mind” and so all kind of observation device even not living is a “mind”?
Zen Arado: aren’t you just saying that we impose frameworks onto the world i.e the classical model and the quantum model but neither solve all the problems?
Wester Kiranov: and it would also be intersting to look at how the feedback loop that make living things living might differ from the feedbackloops that make observations
Zen Arado: so we just kep loking for better models?
Mitsu Ishii: Yes, basically that’s correct. One could think of our idea is an attempt to apply QM to Bateson’s ideas, which were entirely classical.
Mitsu Ishii: Gilles: yes, the idea is any feedback loop with the right properties could be called a mind. But in keeping with Everett, the idea is the observation is only relative to that feedback loop. It doesn’t objectively collapse the wave function (but again you can have intersubjective agreement between feedback loops to create an apparent classical world)
Gilles Kuhn: too a reverso you can say that supperposed state are only models inside a theory and have no ontological existence
Mitsu Ishii: Zen: I’m not quite sure what you mean by impose frameworks, but the ultimate idea we have is that in some sense the “real” world may not be said to have anything in it at all, in the sense of objects, etc.
Mitsu Ishii: This is actually something Max Tegmark noticed: that one could think of the universal wavefunction as having next to no information in it
Gilles Kuhn: the wave function in that case is only a fiction usefull to predict statistically some observable things
Katharine Kozlowski is Offline
Mitsu Ishii: so all objects, etc., the whole complexity of the world, exists only relative to feedback loops.
Gilles Kuhn: but in this view mitsu are all the feedbck loop agreeing always?
Mitsu Ishii: it’s as though there’s a featureless block of “ground of being” with nothing in it, but there are a vast array of possible worlds which exist in that block of ground relative to each other.
Mitsu Ishii: In this view there are vast numbers of possible universes, just like Everett. with many different communities of feedback loops
Gilles Kuhn: or your interpretation need to incorporate multiple universe a la everett?
Mitsu Ishii: it’s more akin to the many-minds idea as expressed by Lee Smolin for example.
Mitsu Ishii: the same “ground” could support many possible communities of feedback loops from this point of view.
Mitsu Ishii: a “community” is a set of feedback loops which have intersubjective agreement
Birric Forcella: No, if I understand it right, then the superposed world is the real world and what we see in contradistinction is due to the feedback loopiness of our minds.
Gilles Kuhn: another problem i see these feedback loops are themselves object no?
Zen Arado: don’t feedback loops always have a set point value?
Wester Kiranov: isn’t intersubjective agreement just a bigger feedback loop?
Mitsu Ishii: yes, you could say the intersubjective agreement is just a larger feedback loop, precisely.
Zen Arado: and are negative and positive?
Gilles Kuhn: and so as object have their own wave function
Birric Forcella: I agree about superposition being the real state of the universe, but you are not explaining that from the moment of the wave collapse the superpositions take a different shape. So the collapse or something is not entirely an artifact of perception
Mitsu Ishii: well, as to what the feedback loops are, that’s the part we haven’t worked out
Birric Forcella: The many worlds do remove that problem.
Mitsu Ishii: how they would actually work in some specific model is something that we just started to think about, but haven’t worked out.
Gilles Kuhn: yes but in a very costly way ontically speaking birric
Mitsu Ishii: The problem with any non-many-worlds interpretation is that it actually creates a severe difficulty when you start to work out the implications.
Mitsu Ishii: basically due to the schrodinger’s cat problem
Mitsu Ishii: you end up with endless infinite regresses.
Mitsu Ishii: for example, consider a quantum computer
Gilles Kuhn: too mitsu may i stress that if you take qm in a non realist vision of science these problems are less big you can see these aparent paaradox as technicalities that dont pose problem as to the fact qm is technically efficient
Mitsu Ishii: I would say my interpretation is sympathetic to anti-realism but not entirely anti-realist.
Mitsu Ishii: that is to say, it would assert that nearly everything we observe doesn’t have “objective” existence aside from perception, but
Gilles Kuhn: indeed your interpretation create a full new ontological universe based enterely on batesonian minds
Mitsu Ishii: it does assert there is some sort of “ground” which supports it all, which has some sort of properties.
Mitsu Ishii: and those properties would allow the batesonian feedback loops to come into relative relation to apparent objects and so on.
Gilles Kuhn: i must say that its one of the most elegant presentation of radical idealism i have seen
Mitsu Ishii: haha
Mitsu Ishii: it is akin to radical idealism but since it posits some sort of ground, it’s not as though one could just “wish away a rock” or anything of that sort.
Gilles Kuhn: but the feedback loops themselves are object and so they are subjected to wave collapse too…
Mitsu Ishii: the properties of the ground, whatever those are, would constrain what sorts of things one could do relative to the universe.
Gilles Kuhn: indeed mitsu but its very close of tibetan view of the world as projection of our minds
Mitsu Ishii: yes, that is correct, Gilles. As Wester pointed out, that would just be a larger feedback loop, in a sense.
Mitsu Ishii: It’s certainly compatible with the Buddhist notion of co-dependent arising, in which mind and world co-arise.
Gilles Kuhn: now the big problem from an empiricist point of view is how to test this idea of yours
Mitsu Ishii: It would provide the beginning of a possible research project which would present a physical theory for that.
Gilles Kuhn: or in others terms can you formulate prediction that are different from standart qm
Mitsu Ishii: Well, there are ways of testing Everett, from what I have read. If you can do a computation and then “erase” it in a quantum computer, that would suggest that wave function collapse is not objective.
Mitsu Ishii: If you can test Everett then you could test this idea.
Mitsu Ishii: Penrose, for example, suggests that wave function collapse is objective. If that is true then my idea is wrong.
Wester Kiranov: i think it might be possible to think of nonstandard feedback loops. in principle. they might ’see’ something different.
Mitsu Ishii: I believe the recent experiment showing macroscopic object superpostion tends to argue in favor of this approach.
Mitsu Ishii: Yes, I think you are correct, Wester. There may be room for perception beyond the normal classical perception.
Mitsu Ishii: It would be an interesting mathematical question if we could formulate this idea more precisely mathematically.
Gilles Kuhn: well an entanglement cease to exist if there is interaction with other systems the ceasing to exist is like a wave collapse even if the others sytems afre not observers
Mitsu Ishii: decoherence remains a limit in this approach. Once a system interacts with the world enough to decohere it, then superposition is lost, effectively, even if Everett is correct.
Mitsu Ishii: I have the idea that there may be a way to think about gravity from the perspective of this interpretation.
Gilles Kuhn: exactly and as our world is full of particle etc to maintain a system in entangled state is not easy at all
Mitsu Ishii: However, that’s a long story in itself which I will save for another time
Mitsu Ishii: If my vague intuition about that is correct, then gravity might turn out to be another way of testing this.
Zen Arado: ty for lecture gtg bye
Gilles Kuhn: well this was fascinating may i propose to continue this discussion next week too if you can be here mitsu ?
Agatha Macbeth: Bye Zen
Mitsu Ishii: it might be related to the “entropic gravity” idea that recently came out by Verlinde. I think that’s a very exciting idea
Mitsu Ishii: sure, I’ll put it on my calendar.
Mitsu Ishii: to make sure to come next week
Agatha Macbeth: Yay
Gilles Kuhn: very very nice
Mitsu Ishii: thanks everyone!
Bleu Oleander: ty!
Gilles Kuhn: thanks to you mitsu
Agatha Macbeth: Thank YOU

permalink | comments(rss)

comment trackback