A wave function does not collapse upon detection?

Does a wave function collapse if there is no conscious entity to measure that?

It is a philosophical question, but philosophy is precisely designed for dealing with what is technically unfalsifiable in hard science—the “far lands” of modeling the system from within itself.

Ah, it's done between satellites! Interesting, thanks.

Contrary to what used to be hammered into people, it must be observed by the observer, that is, by you. Any random interaction will not trigger the collapse. An interaction must itself be observed by the observer, otherwise it will not cause a collapse.

It seems that the universe works pretty much as Haskell does. Things that don't matter to anything else do not happen.

There's a difference between strong and weak quantum measurements. Experimentally we can now even apply feedback on quantum states to stabilize them, which requires making partial, non-destructive measurements of a quantum state and then applying signals to correct deviations.

The whole talk about "wave function collapse" originated from the beginnings of quantum mechanics where the measurement process was poorly understood, in my opinion it should be replaced by the modern understanding, in which a measurement is a combination of an entanglement process between the system under test and a large external system, followed by decoherence caused by the many degrees of freedom of the external system. That produces wave form collapse, but it's not an abrupt or absolute process but rather a continuous, controllable one.

This is all "operationally correct".

If you are intrigued as to what it means for the wave function to collapse, and related questions such as at which speed does it collapse, etc, you must consider what it means to measure something.

If you want to know the property of some particle you need to interact with it ... at least with another particle. The system of those two particles is described by a wave function. This wave function evolves according to the Schrödinger equation and it will tell you that after the interaction you have a superposition of two states: one where the particle you're measuring is inside the box and the measurement probe particle is carrying that information, and the other state that tells you that the particle is outside the box *and* the measurement probe tells is carrying that information.

The wave function of the system hasn't really collapsed to a single state.

Now, let's say that your measurement apparatus records the information of this test particle by e.g. printing it out to a piece of paper or whatever (bits in hard disk)

The wave function of the whole system now captures the superposition where the measured particle was in one state and the paper contains black ink in some places and not in other places, superimposed with the other state where the measured particle was in the other state and the paper carries ink in other locations.

Nothing has collapsed yet; you just have a very complicated wave function. By processing the measurement you effectively introduced a "causality amplification" process, but one that operates on the whole of the wave function.

But, you may protest: "I see only one outcome of the measurement!"

The wave function of the universe is in a superposition of your body (and thus brain) being in a state where you have observed ( and recorded such observation) one measurement outcome and another state where you have observed the other outcome.

Both states are equally real and present in the wave function of the universe. But each of the brain states in the superposition doesn't interact with the other states do each future though you have in one of the superimposed states are effectively independent from the others, giving the impression that the wave function has "collapsed".

Some people like to say that the "universe has split" into independent world lines or something like that. Some people find this terminology helpful, some find it confusing. But the thing is, this just follows from applying the Schrödinger equation without making arbitrary exceptions when humans get involved.