Artificial Intelligence, The Hedonic View of Appreciation and the Threat of Appreciative Deskilling

Abstract This article draws attention to some of the goods of our appreciative practices that we risk losing as AI becomes common place. I begin with a critique of what I regard as an unduly narrow conception of appreciation as, roughly, liking something, that is operative in much of the empirical aesthetics literature generally and the empirically oriented AI-aesthetics literature in particular. In contrast, I offer a sketch of appreciation that emphasizes perceptual-cognitive skills. Insofar as we deploy those perceptual-cognitive skills with the aim of tracking the exercise of human agency in the shaping of artworks, AI’s potential to decenter human agency from our art-making practices threatens them. The threat of appreciative deskilling arises because appreciation is attuned to the exercise of human agency; appreciating an artwork involves the apprehension of what a human agent does, broadly speaking, in a particular medium, and what they achieve, relative to their aims, in doing so. It is, therefore, far from clear how appreciation might proceed, let alone how such skills might be brought to bear as part of that activity, if there is no human agency for it to track.