ZapperZ said:
In such a situation, classical mechanics STILL dictate that if you do know all there is to know about the process, then the outcome of that coin toss is deterministic. No such description exists for the QM case. If your friends are saying that there has to be a similar underlying mechanism that causes a selection of states, then they are speculating beyond what is known conventionally.
I would like to add to what Zapperz explained, that we have serious indications that even if there would ever be such a deterministic explanation for the randomness in QM, that this explanation has to satisfy some bizarre properties which make us believe that it is in principle impossible to know the state of these underlying properties.
The reasoning goes as follows: Bell's theorem tells us that such a deterministic explanation (it is called "hidden variable theories") can only coincide with all QM predictions if it is "non-local". Experiments indicate strongly that cases where Bell's theorem plays a role, nature still seems to follow the QM predictions (I say "indicate strongly" because there are some rather implausible, but possible, loopholes in the experimental setup).
But a "non-local" theory is a theory, according to special relativity, in which cause and effect can switch roles (if event A causes event B in a non-local way, then for another observer, B arrives before A arrives). So if we are not going to be able to influence the past, we better never have experimental acces to this underlying mechanism.
But of course, it is always possible that QM is wrong, that special relativity is wrong (and that the moon really is made of green cheese), so there is always a scientific doubt to statements that contain a "never" or an "always".
But for all what is known today, we really think that the randomness in quantum mechanics is fundamental.
cheers,
patrick.