Read it!

Do artificial reinforcement-learning agents matter morally?

Tomasik, Brian. 2014