Effective altruism in the garden of ends — LessWrong