Nonideal universals can resist oppression without identifying alleged chauvinist construction of rights bearing subject

(Daniel Estrada, Ethics@Newark, Human supremacy as posthuman risk. Journal of Sociotechnical Critique, 1(1), 1–40.

https://digitalcommons.odu.edu/cgi/viewcontent.cgi?article=1013&context=sociotechnicalcritique)

AI ethicists who are sympathetic to the critique of human supremacy developed in this paper may find themselves facing a dilemma similar to the one Khader describes in the transnational feminist discourse. Like missionary feminism, human supremacy continues and expands a Western imperialist ideology that some ethicists will be hesitant to endorse. However, abandoning humanist ideals may seem to leave us in a moral vacuum, without conceptual guidance for protecting the interests of persons and communities under genuine threat of systemic oppression. This concern might simply reflect a parochial Western bias; as mentioned earlier, many indigenous or non-Western perspectives operate on nonanthropocentric ethical frameworks. Nevertheless, as a practical matter, the international law and policy framework of human rights is not just abstract, ideal theory; it is also an established institutional reality, one which can mobilize material resources and direct international cooperation in the service of real human interests. This is not to say the framework of human rights achieves every ideal it aspires to, but simply that it is an important and useful tool in the toolbox, one which would work better if it were applied more consistently and inclusively, rather than if it were whittled down or eliminated, at least absent better options on the table. If the critique of human supremacy developed in this paper implies that the framework of human rights is an imperialist project, some AI ethicists may be tempted to bite the bullet and admit their preference for an imperialist ethical regime. However, Khader’s discussion of nonideal universalism suggests a way for AI ethicists to avoid the apparent compromise of principle, where we can productively engage an institutional framework like human rights while rejecting the imperialist framings of reactionary posthumanism and human supremacy. Following Khader’s nonideal universalism, we might construct a nonideal AI ethics that begins in conditions of systemic oppression and seeks to reduce or eliminate that oppression. This project does not require a monolithic or exclusive idealization of human agency or cognitive capacity. We can use the framework of human rights to protect the interests of actual human communities without treating that framework as a chauvinist characterization of the ideal rights-bearing agent, one which might justify the exclusion, abjection, or oppression of other humans or nonhumans. A nonideal AI ethics rejects the notion of a single ethicscompatible cultural form of agency or capacity, thereby undermining the idea that an idealized Western anthropocentrism is the ethical solution. This opens space for rethinking the sociotechnical matrix of human and artificial agents in dynamic political terms, without being misdirected by an impulse to comparison, exclusion, or hierarchy. Robots are situated within existing sociopolitical structures to materially extend and reinforce systems of domination and control, but they also encounter these systems as a practical constraint on their operation, and so can also be mobilized to resist, dismantle, and repurpose these systems. As such, robots may already figure within normative communities as having varying degrees of agency, complex social alliances, and relationships with other human and nonhuman agents. A critical discourse on robots and AI in nonideal conditions can recognize how overlapping structures of institutional oppression situate robots as both agents and targets of power—as agents whose identity and operation must be made available for inspection, public scrutiny, and abuse (Romero, 2018; Smith & Zeller, 2017; Brscić et al., 2015; Salvini et al., 2010). To some extent, the actual circumstances of these arrangements and the moral vernaculars of the communities involved should play a role in our collective assessment of the robot’s impact and value.