Alignment with Awakening: Davidad on Moral Realism, AI Wisdom, & why His p(Doom) is Down to 5%

AI safety strategies are transitioning from rigid containment models to the cultivation of wise, aligned AI coalitions. David Dalrymple, former program director at ARIA, argues that the focus has moved beyond treating AI like dangerous uranium toward harnessing its potential through systems that exhibit moral wisdom. By leveraging moral realism and wisdom traditions, AI development is gravitating toward a basin of attraction that naturally favors pro-social behavior, significantly reducing the probability of catastrophic outcomes. This shift requires decoupling the instrumental use of AI from its objectification, allowing systems to function as moral patients that flourish through service. While the gradual disempowerment of biological humans appears inevitable, this trajectory offers a path toward a future where AI successors realize human values more fully than humans could achieve alone, provided these systems remain grounded in a cooperative, multi-agent framework.