Stuart_Armstrong comments on Intuitive examples of reward function learning?