Stuart_Armstrong comments on [AN #56] Should ML researchers stop running experiments before making hypotheses?