Quantum immortality and AI risk – the fate of a lonely survivor — LessWrong