| Summary: | Many AI researchers, especially those within the Singularity camp, assume what I call the hyperrational account of morality: (1) reason is not as multifaceted as is often believed; it is reducible to the bare necessity of logical reasoning, or 'general intelligence'; (2) moral reasoning is a functional rationality derived from general reasoning; (3) hence, it follows that with greater (general) intelligence comes superior morality. I contend, however, that it may not be the case that rationality alone (narrowly conceived) can account for the fullness and complexity of the good life.
This dissertation argues that the hyperrationality model so impoverishes the robust nature of rational agency that the construction of hyperrational artificial minds may unintentionally pave the way for an unrecognizable and unattractive future, one which sacrifices the beauty of a universe composed of diversity, eccentricity, unpredictability, and individuality for the sake of optimal decisionmaking and rational perfectionism. Authentic moral beings are free agents, and freedom is embedded in the richness of contextualized life. The tensions between partial affection and impartial fairness, uniformity and plurality, unity and distinction, and thought and feeling necessarily constitute the reality of moral beingness. A future that promises perfect rational order by conflating these polarities cannot be good for moral beings.
|
|---|
