Artificial General Intelligence and Artificial Human Minds: A Functional Convergence Under Epistemic Opacity

Recent discussions in artificial intelligence have drawn a distinction between Artificial General Intelligence (AGI) and an Artificial Human Mind (AHM), most notably in the work of François Chollet (2019). At first glance, this distinction appears to hinge on the contrast between functional problem-solving capacities and the richer phenomenological dimensions of human mentality. While the human mind indeed comprises more than intelligence alone, this paper argues that the phenomenological components that exceed intelligence are epistemically inaccessible to others and enter the shared world only through their causal effects on behavior, expression, and interaction. Crucially, this epistemic opacity is not unique to artificial agents—it already characterizes human-human and human-animal relations. Human beings have always navigated this opacity through inference, expression, and practical engagement rather than through direct phenomenological access. This paper introduces the concept of Artificial Social Mindedness (ASM)—the threshold at which an artificial system satisfies all functional requirements for participation in human social life, regardless of phenomenology. The central argument is that ASM, not AHM, is the relevant criterion for practical coexistence with artificial agents—and that the principles grounding this criterion are already operative in how humans relate to one another and to non-human animals.