Giving the results of scientific research away — LessWrong