Could a physicist please clarify this?
When does "collapse" happen in principle? There's no clear answer. Yet, when you observe a system yourself, you are free to say it has one result (at least, in "your world / branch" if we use MWI lingo). So there is (possibly) a sense in which you are the only observer -- at least, w.r.t. the place you think of as "the world."
(And of course when we say "you observe," we mean "you experience". If only your toe measures it, you may treat your toe as an external measuring device. This is how "consciousness" enters the picture, though nobody agrees on what exactly it means.)
But from the viewpoint of the toe would it have collapsed?
Extrapolating that, why isn't your consciousness just another measuring device that's become entangled with the system?
Similarly, your consciousness could be considered just another measuring device from the perspective of someone outside the system. But by virtue of being "outside the system," their results have no bearing on you either.
You are the point at which "the magic happens" in your world -- which is just the place you normally call "the" world. It is the point at which neither you nor anyone else you can communicate with (i.e., no one else in "the world") can demonstrate interference, even in principle.
It does not happen at all. There is no measurable distinction between a particle whose wave function has collapsed and one that hasn't. If there were a discernable difference the "spooky action at a distance" when an entangled pair is collapsed could be used for faster than light communication.
>> When one system measures another, there is no collapse, only entanglement.
This is the first time I've seen this explanation. To me it feels like an attempt at resolving the issue, but I think not.
The right answer AFAICT is that we must embrace non-locality. But that's still another sort of punt.
Also, this surprises me (assuming you studied QM formally):
> This is the first time I've seen this explanation.
Entanglement is basically the definition of measurement (in this context).
> As a non-physicist who only knows QM from popular science texts, I always thought that wave function collapse does not require any human intervention and that this can be shown.
When pop sci treatments explain that any physical system interacting with the apparatus "measures" it, they are just talking about what you call "pre-measurement." This causes a lot of confusion for students.
Suppose you are making a universe. You want to ensure that the universe you make is interesting, i.e. contains consciousness. So you set up the rules like QM: everything exists in probabilistic superposition until a consciousness observes it. One of the superpositions contains a consciousness, and that collapses the wavefunction guaranteeing consciousness in your universe. It's a way of using survivor bias in your favor.
Disclaimer: not to be taken seriously.
Disclaimer #2: Also, coincidentally, not related to poster drdeca. Unless he is me from another superposition and we haven't collapsed yet because you all are not as conscious as you think.
I suspect one could think of QBism as a theory of rational belief for agents in a context where details of quantum mechanics are highly relevant.
Just as bayesianism is a theory of rational belief more generally.
Perhaps we are using “theory of” in different ways?
How do you know?
I agree! However:
> because the vast majority of data we have excludes it
No it doesn't. In fact, there is not a single experimental result that rules this out and I doubt one could ever devise such an experiment in the first place.
All experimental results are consistent with the observer modeling the universe with the Schrödinger equation (i.e. with unitary time evolution / without collapse) and assuming that a measurement by said observer makes the wave function collapse to the appropriate eigenstate.
What experiment shows this?
As you know, in the classical double-slit experiment, the screen is the only detector, and launching particles even one at a time through the slits produces the interference pattern.
However, if you add a second detector in one of the slits capable of measuring if a particle passes through that slit or not, the interference pattern on the screen disappears, regardless of how many particles you send through.
This is normally taken to mean that the slit detector collapses the wave function (in Copenhagen parlance); or that one particular state of the particle becomes entangled with the initial detector, which is already entangled with the screen, to there is no more constructive interference to be seen (in Many Worlds terms).
Here is a more complete explanation:
https://phys.org/news/2011-01-which-way-detector-mystery-dou...
The delayed choice quantum eraser is a much more complex experiment.
From what I've read, the existence of "collapse" of the wave-function is not established, and not even widely accepted as more than a useful computational tool by practicing physicists.
Full disclosure that I’ve never really been able to pin down the QBist notion of an agent myself and do not find QBism (or any such “instrumentalist” interpretations” of QM) appealing.
Do you have a reference for that?
Long story short, QBism inherits "agent" from Bayesian decision theory. In that context, we talk about gamblers making bets on the outcomes of dice rolls. In QBism, we talk about scientists making bets on the outcomes of measurements. What exactly can be a "gambler?" Can a dog gamble? What about a trained dog? Or a flea? The answer is: if the shoe fits, wear it. Same goes for QBism. If a dog walks into a lab, lights up a pipe, scribbles equations on the blackboard and starts aligning beamsplitters, then QBism regards him as an agent.
A personalistic interpretation requires a "person". Would a rock or a cell phone qualify as a person? I'm not even sure the authors of that reference would accept a dog, despite saying the concept is extremely flexible (they discuss agents "social" agents but not "dumb" agents).
It's not clear to me that the mathematical description can exist independently of the agent that has some knowledge, for some definition of agent. Even if it can be make to work with no-one's "abstract knowledge" what would be the physical meaning of that? (Not that the physical meaning when agents are present is clear, to be fair.)
Take the electron double-slit experiment. Suppose no human checks which slit the electron goes through, but the "which-path" information leaks out into the environment somehow. The interference fringes will be destroyed anyway (called "environmental decoherence").
This is usually framed by saying "oh look, the environment measured the system". In fact, what happened is that (1) you prepared what you believed to be an electron beam in a coherent superposition; (2) you left it open to the environment; (3) you used quantum theory to model the system-plus-environment and predicted that you would not see interference fringes when you check; (4) YOU observed the dots on the glass plate and confirmed that there were no interference fringes. The whole story is told in terms of what YOU believed, expected, and finally what you saw. That is QBism's whole point.
Put it another way: in QBism wavefunctions just ARE sets of probability assignments to certain actual events that could be observed. And probabilities are human constructions: we invented them to help us reason about our expectations. So, without humans (or other beings capable of reasoning about expectations), there are no probabilities, hence no wavefunctions either. But stuff still happens, even when there's nobody around to see it or assign probabilities to it.
The key conceptual leap is separating "wavefunction" completely from "reality". There's more to quantum physics than just the wavefunctions -- the reality lies in the events that happen, not in the quantum state itself.
It is, in principle, impossible to determine this, because eventually humans will see the output of the experiments. However, we can say that if humans are special, the only things able to cause “measurement”, then our specialness also gives us time travel powers that can only be used for “measurement”, and Occam's Razor says that's probably nonsense.
Why? Because the Quantum Eraser Experiment comes out as it does?
We don't have an accepted theory on what "measurement" is in Quantum Mechanics yet.