Can humans and AI robots be friends?

摘要

In the divide between ‘Robophobia’ (the majority position) and ‘Robophilia’ (minoritarian), the latter alone allows for the possibility of friendship between them. This chapter first reviews the four main barriers adduced to preclude it: Normativity, Emotionality, Consciousness, and the absence of Qualia. All barriers accentuate deficits on the part of AI entities/beings and ignore their capacities for (i) continuous learning, (ii) error correction, and (iii) self-adaptation of their uploaded skills. My case for the possibility of their friendship is based upon the emergent relational consequences of synergy when the two work together. On the Critical Realist causal criterion their friendship is freed from essentialism, exceptionalism and speciesism. Friendship between them entails ‘joint action’ but not ‘joint commitment’, as illustrated by a thought experiment in medical research. From synergy other constitutive emergents are ‘trust’ and multi-dimensional relationality, capable of enhancing AIs current work in care of the elderly and autistic children