Superpositions with two observers This is a bit of an odd question. I'm not a physicist, so bear with me if I say something wrong.
Lets say you have some sort of quantum event where matter is in a superposition. Standing next to you is another scientist waiting to observe the results (and, in theory, collapse the suposition). You go to get a cup of coffee while your fellow scientist stays in the room. Your fellow scientist observes the result of the experiment while you are out of the room but does not tell you what the result was. There are three possible options here for when you return to the room:


*

*The matter is still in a superposition for both you and the other scientist.

*The matter is no longer in a superposition for either of you (even though you have not observed the event and have no knowledge of what happened).

*The matter is in a superposition only to you.


So which option is it? If the answer is option 2, would that mean that we as humans have some kind of "superpower" forcing quantum superpositions to collapse as soon as we see them?
 A: This problem is called Wigner's friend. Basically the answer according to quantum mechanics is 2. If my friend makes a measurement causing collapse of the superposition, then when I return I will not see any quantum interference, so a superposition state would predict the wrong outcomes. If my friend doesn't tell me what happened in his measurement, then I must assign a mixed state, or classical probability distribution, to the system. 
Whether you think this is weird or not depends on your idea about what quantum mechanics is telling you. If you think that quantum states just represent knowledge, then there is nothing weird: collapse just means updating your probability distribution when you learn new information. If you think that quantum states are physically real things, then there is no problem if the system is localised: the measurement must involve a physical interaction which changes the physical state. 
The spanner in the works occurs when you deal with entangled states. Then the collapse in my lab also appears to have consequences for measurement outcomes in distant places. Bell's theorem tells you that you cannot have both a local and realistic description of this process. If you think quantum states are only information, then you must accept that physical properties of quantum systems do not exist independently of observation. If you think quantum states are physical objects, then you must accept that the interaction in my lab had a non-local influence on the system in the distant lab. Most people like locality, so choose the former option. Particularly because the concept of locality is deeply rooted in our understanding of elementary particle physics and gravity. But the two options are not distinguishable by any currently known experiment, so the inevitable and continuous arguments between people from the two camps are pretty pointless. Nevertheless, I expect at least one person will start arguing with me over this post :)
