Schrödinger’s Cat When Nobody Is Looking
nautil.usOne of the problems is that quantum theory is obscure (to say the least) regarding what it claims about the nature of the world when no one is looking. Is the involvement of a consciousness required for the theory to make sense, and if so, does that include a mouse’s or a fly’s? In particular, the specification of what constitutes a measurement is irreparably vague. Perhaps all that’s needed is a large enough apparatus. But what’s large enough? And what happens at the boundary? These issues are referred to as the measurement problem.
In my utterly uninformed imagination I've reconciled this by defining observation as the point where the observer (whether it be a simple apparatus like the sensor on a photon counter, or a complex and squishy one like my consciousness) becomes "intertwined" with the observation. i.e. Once the information has reached some particle, that particle's view of the waveform/event has collapsed.
It's easy to forget the actual conveyance of the information has some physical manifestation (photons hitting my retina, electrons bumping against transistors in an IC and eventually manifesting as dots on a screen which send out those photons) and I picture those causal steps as propogating constraints through the universe. Light speed sets a maximal boundary on how much of the universe may be "intertwined" with the observable, but even if the data is sitting on the computer next to me, until I take a peek (or otherwise subject myself to any consequences stemming from them) those constraints haven't yet permeated any of the bits of the universe that make up me.
Akin to how relativity permits two different observers to hold differing views of some phenomena (eg. silmultanaety) this worldview allows me to imagine the cat is both dead and alive even though my computer or the Geiger counter may know the correct answer.
I'd love to hear from real quantum physicists whether this interpretation is bunk or has some validity (and if so, whether someone else arrived at it before me and gave the theory a name).
I'm not a "real quantum physicist" but I've been studying QM as a hobby for thirty years. Your intuitions are not far from the mark. What you call "becoming intertwined" is a real physical phenomenon called entanglement. The problem with your definition is that there is no "point" at which entanglement "happens". The creation of entanglement(s) is a continuous process, not a discrete one, so the transition from "unmeasured" to "measured" on this view is also continuous. This process in the aggregate is called "decoherence" and it can be described very precisely, and it correctly predicts all of our observations, including the fact that all macroscopic observers will agree that either the cat is alive, or that the cat is dead.
The problem is that nowhere does the math say that the cat actually is alive or dead. Again, mathematically this is a consequence of the fact that the Schroedinger equation is linear, so if you start with a superposition of states, that can only evolve into another superposition of states. The math says unambiguously that the cat and all of its observers are in superpositions. In order to extract a single privileged objective classical reality you have to add some additional mathematical constraint, and no one has been able to figure out a way to do that that isn't either ad hoc or at odds with the data.
And it gets even worse than that because of the Bell inequalities, but that's a whole 'nuther kettle o' worms.
Or. Or you can go with poorly-named "many worlds" by applying the Occam's razor to the ridiculous concepts of "classical observer" and "wave-function collapse".
There is no such special thing as a "classic observer": all instruments and people are in the end made out of quanta, and are themselves huge quantum systems.
There is no such special thing as a "wave function collapse from observation": what happens is that the observer (a quantum system) becomes entangled with the observed system. That's it. Initial state of "[ignorant observer] * ([cat is dead] + [cat is alive]) / sqrt(2)" which is equal to "([ignorant observer, cat is dead] + [ignorant observer, cat is alive]) / sqrt(2)" evolves into "([observer thinking of a dead cat, cat is dead] + [observer thinking of a living cat, cat is alive]) / sqrt(2)" without breaking the unitarity.
Yes, for "the" observer, her being part of this state, it looks like "the collapse has happened". That's how you get "many worlds": the state of the universe is a huge sum of independent states that continue to evolve independently, each of them evolving further into a sum of independent states.
IANAQP, but your view comes close to Rovelli's Relational QM. It may not be entirely precise to say that your computer knows the "correct" answer -- it just know its answer. Yours still does not exist. Once you know your answer, its answer "becomes" correct (because you must agree).
(Of course, given that the very existence of the Geiger counter or computer may depend on which branch occurred (perhaps there was a bomb on one path?), it's unclear whether you are free to think of there even being an "it" that could have an answer, either.)
> the very existence of the Geiger counter or computer may depend on which branch occurred (perhaps there was a bomb on one path?)
At the quantum level, the quantum degrees of freedom in the Geiger counter or the computer still exist even if a bomb goes off. They just have a different relationship in terms of interactions than they would if the bomb didn't go off.
Sure, something has recorded the state. It's just that we like to think of there being fixed, discrete entities that have knowledge, whereas really it's more like that knowledge is an integral part of their identity.
I sympathize with your view a lot and it is, in some way, also the way I (degree in mathematical & theoretical physics) think about it.
Do note, however, that the wavefunction and its collapse are at odds with special relativity and thus cannot be an object of reality. (Provided you're interested in realism to begin with.) Take two entangled electrons, for instance, which get sent out from the origin 0 in opposite directions towards detectors A and B, respectively, which are placed at equal distance from 0.
Now take two observers, one moving in the direction from 0 to B and one moving in the opposite direction. Then, depending on which observer's POV you take, it is either detector A might makes the wavefunction of the two electrons collapse (because the electron 1 reaches detector A and interacts with it before electron 2 reaches detector B) in which case interaction of electron 2 and detector B doesn't do anything. Or, vice versa, it is detector B which makes the wavefunction collapse but then detector A doesn't do anything special.
Either way, if the collapse were actually something physical, both observers would have to agree on the causality chain of events. But they can't. In fact, varying the above setup a bit, they won't even in general agree on whether the wave function of a given system has or has not yet collapsed.
What I particularly find interesting is that there is a nice correspondence of this in programming language theory where depending on the evaluation strategy the same values are derived differently from a set of expressions i.e. values are computed in different orders and different causality relationships arise (causality in the sense of which expressions cause another expressions to get evaluated e.g. call by value vs call by name) but ultimately you arrive at the same result (or varying the initial “setup” may result in some expression not being evaluated)
Thanks for weighing in, I really appreciate the comments people are leaving. One thing I've struggled for clarity on: in examples like that one, is "agree" meant to apply at the instant either observes their event (ie. some omnipresent referee not subject to relatavistic limitations) or later on when they get together to compare notes?
Everett proposed relativistic interpretation. The measurement happens solely due to physical interaction and is properly contained in light cone. The difference is that the cat has two states: one dead and one alive, when you physically interact with them, your state splits into two, each state sees the respective cat's state, both are real, but the picture is relative. The observation corresponds to not what exists, but to what physical interactions happen, and because these split states don't interact, they don't see each other as if it was a collapse.
I'm far from being an expert, but does your interpretation allow for quantum computation to work? Sounds neat if it does.
This is a good new theory which makes SENSE as far as I can understand.
But, if we apply Occam's Razor then isn't the explanation to all quantum weirdness simply that we live in a simulated universe?
So, I am missing something, how does it work with the double slit experiment ? the result should never act as a wave without observation if we have spontaneous collapse all the time ? Or is this spontaneous collapse quite rare in practice ?
Yes, that's right. Extremely rare. The only reason that the theory works at all is that a single spontaneous collapse can propagate throughout a complex system of entanglements and collapse the whole system.
The nice thing about spontaneous collapse is that it makes a testable prediction: there should be a scale at which the behavior of an isolated system starts to show divergence from quantum predictions. So far that prediction has failed to be demonstrated, but people are still working on it.
> is this spontaneous collapse quite rare in practice ?
It's extremely rare for a simple quantum system like a single electron; but it is happening basically all the time for a very large system like the detector in the double slit experiment. So basically, the electron is virtually certain to get all the way through the double slit experiment without any spontaneous collapse, but as soon as it interacts with the detector screen at the end of the experiment it will have to collapse basically immediately, because the detector screen is always having spontaneous collapse events and the electron is now entangled with the screen and has to collapse along with it.
This seems like a reinvention of the Penrose Interpretation.
Both are collapse theories [1][2], but presumably not identical being that the author is aware of Penrose's work and credits Penrose earlier in the article for some terminology as well as acknowledging his influence on the theory in the following footnote form the article:
> 6. In this, we have been strongly influenced by considerations in this regard made over 3 decades ago by works such as Penrose, R. Time asymmetry and quantum gravity. In Isham, C.J., Penrose, R., & Sciama, D.W. (Eds.) Quantum Gravity II (1981); Wald, R.M. Quantum gravity and time reversibility. Physical Review D 21, 2742 (1980).
The first footnote tells us that the author is referring to GWR theory [3] specifically (which is distinct from Penrose's):
> 1. Ghirardi, G.C., Rimini, A., & Weber, T. Unified dynamics for microscopic and macroscopic systems. Physical Review D 34, 470-491 (1986); Pearle, P. Combining stochastic dynamical state-vector reduction with spontaneous localization. Physical Review A 39, 2277-2289 (1989); for a relatively recent review see Bassi, A. & Ghirardi, G. Dynamical reduction models. Physics Reports 379, 257-426 (2003).
1. https://en.wikipedia.org/wiki/Objective-collapse_theory
2. https://plato.stanford.edu/entries/qm-collapse/
3. https://en.wikipedia.org/wiki/Ghirardi%E2%80%93Rimini%E2%80%...
This article contains a consise explanation of the black hole information paradox, and, for the first time, I feel as though I understand what the issue is (assuming that it is not glossing over some important issues.)