Is the collapse of a wave function deterministic or random?

[.Scott]

But if there is no collapse, then the different eigenstates superimpose and produce an interference pattern (say, of a double-slits experiment). If the wave function collapses to a single eigenstate, then the interference pattern is destroyed. So the final result is different.

It would mean that you're in a superposition of conscious states. How would you know that? Is there any experiment that can differentiate when the collapse occurs? The apparent lack of self-eating cat suggests that many possibilities are lost as the wave function involves more and more of the macroscopic world. But even that does not suggest that a full collapse ever happens.