Why Many-Worlds Is Not The Rationally Favored Interpretation — AI Alignment Forum