Errors in the Bostrom/Kulczycki Simulation Arguments — LessWrong