Abstract

Is phenomenal consciousness a necessary requirement for the creation of artificial general intelligence (AGI)? Despite its philosophical, scientific, ethical and technological significance, this question has received little systematic attention. This paper offers a structured review of current positions on the matter, aiming to clarify the conceptual landscape and identify key points of contention. We begin by examining the dissociation view between phenomenal consciousness and general intelligence, according to which subjective experience and human-level general intelligence are both conceptually and empirically separable, suggesting no necessary connection between them. We then contrast this with the association view, which suggests that phenomenal consciousness plays an indispensable role in enabling the flexibility, autonomy, and meaningfulness that characterize human-level intelligence. This view can be further divided into two groups. According to the functionalist perspectives, phenomenal consciousness is identical to some functional property needed for general intelligence (e.g., global information broadcasting). According to the phenomenological perspectives, phenomenal consciousness is, in itself, constitutive of some key aspect of general intelligence, e.g., genuine semantic understanding or self-concerned adaptive behavior. Finally, we outline future research directions, highlighting key unresolved questions: whether true general intelligence can be fully defined in functional terms; whether phenomenal consciousness has any causal role in general-purpose intelligent behavior; to what extent general intelligence and functional consciousness overlap; and to what extent functional and phenomenal consciousness can be conceptually and empirically dissociated. Rather than taking a position, this paper aims to map the theoretical space and identify the foundational questions that future work must confront.