Will AI superintelligence be evil?

The question might be not as to if or when General AI successfully transcends human intellect, but rather – what is it about the culture, competition and commercial (or, to be frank, geostrategic and complex adversarial) contexts in which these technologies currently effervesce and thrive that might only ever guarantee that they are shaped from, as and towards the replication of the darker angels of our nature?

If exceptionally competitive, predatorial, territorial and ruthlessly pragmatic ends shape the accelerated evolution of these systems, we might not be so surprised to find these qualities starkly reflected back at us from the technologies. Indeed, even where and when collaborative benevolence and ethical forbearance is an authentic, intended and achievable goal – can we ever hope to extricate this from the underlying substrate and gravitational well of blind tribal or reflexive psychological self-interest that we as a species (and Global civilisation) are still aspiring to transcend?