Keith and Andrew discuss the importance of quantitatively assessing AI fairness and ethics, emphasizing the need for human judgment alongside basic rules. They highlight the complexity of developing systems that can engage in conventional agreements about meanings in legal contexts.
