Experience, Identity and Moral Agency in the Age of Artificial Intelligence

AI systems transform all sectors of society, and the ramifications of this revolution in data processing, cognition and learning, communication and social interactions are unforeseeable. For some, AI systems increase human freedom; for others, they threaten the status of human beings: machines have better memory, more efficient strategies to pursue ends, with better means, and they are as predictable as they are capable of learning from mistakes. They may not show the same susceptibility to violence, and thus may even solve the problem of evil that has haunted the history of human morality. I examine why human vulnerability in the form of openness to others, together with the unpredictability of interactions, are necessary elements of moral identity and agency. They can only be overcome at the price of human freedom. Without human freedom, however, interactions are reduced to the exchange of information, needs and desires, and the pursuit of ends that undermine the self-fulfillment associated with moral identity as well as the responsibility that arises from the claims agents make on one another. AI systems have no way to mirror relationships of recognition and responsibility, because they cannot reflect the reciprocal normative claims entailed in moral interactions.