"Quantum Stupefaction" pwns Many-Worlds!
First, let me explain "quantum suicide" (c.f. quantum immortality, check Wikipedia. Later, I explain something even more perplexing.) If we imagine that only your conscious selves serve as possible carriers of what "you observe" as you split up into "many worlds", then you can do this: Set up the Schrödinger's cat experiment with yourself inside. That means, a radioactive nucleus decaying will cause HCN to flood the chamber, so the decay and you dying from it is a chance event and "not determined" by classical physics. In the many-worlds theory, each moment means the splitting of the universe into other universes representing possible outcomes (such as did or did not decay.) Robert Anton Wilson wrote about Everett-Wheeler MW theory, among other things, in the speculative/fabulist fiction work Schrodinger's Cat Trilogy.
Well, there is always a chance you will survive even after many half-lives of the death-dealing source. Hence there are always some "yous" which can enter some "branches" of MW. Sure, "you" die most of the time but the conscious survivors are the ones saying "Hey, here I still am after all this time." Hence you can never die! If you get into the chamber, you will always "still be there" wondering WTH happened and how you survived millions of half-lives of an unstable nucleus. You can even walk out later, by having a temporary period for the nucleus to be allowed to affect the death-dealing device.
BTW this does not depend on whether consciousness collapses WFs or is "weird" since any sentient being can only notice its survival, right? It is eerily reminiscent of the "multiverse" (that has differing physical constants like alpha) idea of explaining anthropic fine-tuning. However, I think that outcome ruins the whole "proportional" argument that sophistrizing MW advocates put out to imply that each of us has the same "expectation" of outcomes in MW as in ordinary collapse. (BTW I believe "decoherence" is a failed, circular argument for making "apparent" collapse work out OK. Note that use of "apparent" yadda is a warning sign of post-modern psychobabble.)
Note, it is quite possible to build this device. I challenge any supporter of MW to have the balls/etc. to actually get in one. We can even arrange a cash prize waiting outside of $10,000,000 or so, that you can only collect when the temporary danger period is over. Since we might set up a chance of only say 1:10,000,000 of a "you" getting out alive after the danger period, it is an eminently reasonable bet for the rest of us. But since you expect to experience surviving and winning, why not show your "faith" in MW and come out a multi-millionaire to boot? Not only that, but you can see the amazed expressions of the versions of us that see you walk out alive! We'd never live it down, heh.
Well, that's already rather weird, but let's take it further into the twilight zone: Suppose we change the quantum suicide scheme so it doesn't actually kill you if the agent nucleus decays. Let's say, it injects you with something causing total but temporary unconsciousness. Let's call this "quantum stupefaction" ;-) Well, the logic regarding "awareness" should work the same way at first. "I" expect to be one of the tiny minority that remains awake and knows it, for how can that be "experienced" any differently than if most of me actually died?
Yet after awhile, all the other me`s start waking up and there are so many more of them then those who never got put to sleep. So, now what? If "I" can "expect" to be able to say, "How fortunate, I didn't even have to become unconscious" what then takes over the expectations when the others wake up? So what do I really have to "likely" look forward to: not even being knocked out at all, or being put asleep and then waking up? The first option follows the original logic of quantum suicide, but the latter follows what we expect from the total chance of being put asleep and then waking up. Quantum [non-]Suicide was at least a potentially viable if wacky notion, but I think this further refinement makes the whole idea (and by extension, many-worlds "theory") look silly.