Introducing Rational Animations — LessWrong