What is the role of the observer? What is the observer? Why does observation cause the wave function to collapse?
For instance say you're performing some slit experiment, directly after you launch the particle there is still a very broad spectrum of possible positions / states, which are all heavily correlated since states that differ only by the position of 1 particle are in some sense "close". However if the particle hits something then suddenly there can be a very large number of particles that have different positions, depending on where the initial particle hit, hence the difference between those states increases immensely, causing them to become decoherent. Hence, from the perspective of the resulting states, there is only a very small region where it could have hit, anything else becomes remotely unlikely.
This interpretation has some rather interesting issues when you try to interpret what consciousness is, but it's the most consistent one I've found so far.
How would you isolate an experiment from the potentially-wave-function-collapsing influence of human consciousness, and yet still produce a usable contribution to science?
There are serious philosophical issues with the idea of an observer, conscious or otherwise, as a useful concept.
Entanglement isn't really the issue, because it doesn't solve the problem. What does it mean to be entangled with a complex state in a state of some form of consciousness?
How much consciousness is needed to make a difference? Why should adult human consciousness be the benchmark, and not semi-consciousness, or distraction, or unconsciousness, or toddler-level pre-verbal awareness, or jellyfish consciousness, or LSD-induced hallucination?
If consciousness is necessary, why doesn't reality stop working when we fall asleep? Why do plants live in a consistent physical reality even though they don't know what a Hamiltonian is?
If physics had code smells, this idea of consciousness might not pass the sniff test.
Of course it's possible it's still the right answer. But if it is, it's interesting there's no useful explanation of anything in the concept yet. Nor is there any formalism to support it. (Obviously there's a formalism for wave function collapse. But so far as I know, there's nothing in the math that says "And these variables are where the observer works his/her/its magic.")
Really it's just something that might be true, maybe, because we have no idea what's going on and it's as good a guess as any other.
The photon is in the state it's in because of the result, which means that, after that the photon enters your eye, your brain is also in the state it's in because of the result. But you could say the same of a dog brain, or a cow brain, or PC with a webcam for that matter. "Consciousness" has nothing to do with it.
Sir Roger Penrose said many times that our current models of Physics are not including Consciousness and that is a big issue. He and Hameroff are working in an interesting line of research that is trying heretically to do just that, model Consciousness pushing the current statu quo https://en.m.wikipedia.org/wiki/Orch-OR
Here is a interview of Hameroff talking about why Consciousness is not an epiphenomenon (as the experiments with paramesiums shows) and is far more than computation: http://youtu.be/YpUVot-4GPM
According to the Many-Worlds Interpretation, it doesn't. Both the thing being observed and the observer exist in many states simultaneously (a superposition of states). Before the observation, those states were independent; the act of observation causes the observer and observed to become entangled, so each state of the observer corelates to a single state of the thing observed. From the observer's points of view, there appears to have been a collapse; but the result of the "collapse" will be different for each of the observer's states.
(Caveat: I am not a physicist, and it's possible the above is not even wrong.)
Though this being HN: Think of it like the edge of the bitcoin block chain two new blocks are can be in an unintended state, but over time the longest one wins.
* Not that this a useful model, but it's much closer than the often repeated voodoo mysticism people spout.
it becomes much clearer (at least to me) once "observation" is replaced with "interaction" because the former is really not possible without at least a quantum of the latter.
It doesn't. Classical behavior is an emergent property of any large system of mutually entangled particles. For macroscopic systems classical behavior is a damn good approximation, but it's only an approximation. See:
http://www.flownet.com/ron/QM.pdf
Or the movie version: