Consequentialism

The key idea of consequentialism is that an action is ethical depending only on the consequences of the act.  This means that the circumstances surrounding the action are irrelevant, only the outcomes.

How is this relevant to AI? Consequentialism would state that an AI system was ethical if the outcomes of its actions are good, no matter how the data was obtained or how the model was trained, etc.  Consequentialism also examines the good for the majority, meaning the outcomes may be suboptimal for some people but as long as the overall good that comes from the use of the model is good, it becomes ethical.

Since this is a discussion for the very first day of class, I’m keeping the assignment very short here.  Please read more on consequentialism below from the Stanford Encyclopedia of Philosophy