The Copenhagen interpretation of quantum mechanics states that a particle is described by a wave function, and privileges measurements by assuming that they collapse the wave function.
Hugh Everett III proposed what has come to be known as the many-worlds interpretation of quantum mechanics, in which measurements are not privileged, and the wave-function never collapses. Instead, the measuring device is part of the wave function, and, at each possible measurement moment, the universe cleaves into multiples, one for each possible outcome of the measurement. Thus, the many-worlds view of Schrödinger's cat is that at every possible decay moment, there is a branching of the universe, into a branch where the cat lives, and a branch where the cat, um, does not live.
This has led to the suggestion that the many-worlds theory could be "proved" via an experiment called quantum suicide, which is a cross between Russian Roulette and Schrödinger's cat. The idea is to sit in a room which is connected to a chamber in which a random radioactive decay either occurs or not, say every ten seconds, with odds of say 50%; if the decay occurs, then a rifle is fired, killing the person sitting in the room. Every ten seconds.
The idea behind quantum suicide is to sit in that room, and not die. Because, after a few minutes have gone by, one's survival would be astronomically unlikely, so one would be forced to conclude that the many-worlds interpretation must be correct. (Which would be comforting, as it implies that consciousness is immortal.)
Our understanding is that there is no evidence that anyone has actually tried this.
