Skynet Meets Planet of the Snakes. Removing metaphysical Impediments to Rogue AI

This paper advances the view that non-rational AI could indeed constitute a significant threat to humankind and that the best way to avert this threat is ethical formation of those involved in AI development. In the light of a novel thought experiment, and by applying an understanding of rationality based on a Thomistic Metaphysics of Participation (TMP) wherein rationality and self-awareness do not emerge, this paper finds no metaphysical impediment to AI posing such a threat in the future. This approach serves to highlight the immediate need for action. The most effective safeguard against a Skynet scenario appears to be ethical formation in justice for the policy and decision makers, and AI developers.