Cognitive Enhancement or Social Atrophy? Generative AI, Neurodiversity, and Human Flourishing

Abstract Debates on the cognitive impact of Generative AI often assume a homogeneous model of human cognition and focus on individual performance. AI’s epistemic authority in contemporary life raises concerns about the social conditions for democratic inquiry and human flourishing. Drawing on Dewey’s pragmatist conception of growth as a collective, pluralist practice and Jaeggi’s critique of forms of life, this paper argues that AI’s tendency toward cognitive homogenization represents a structural risk to democratic deliberation. A critical examination of recent empirical studies on AI and cognition reveals that their experimental designs systematically privilege Neurotypical frameworks, pathologizing alternative cognitive strategies – such as cognitive offloading – while rendering Neurodivergent ways of thinking invisible. Against this, the paper proposes a Disability Justice-informed model of AI design grounded in cross-neurotype solidarity and cognitive pluralism. This paper argues that such approaches overlook both cognitive plurality and the social conditions of human flourishing. From a neurodiversity perspective, Generative AI can function as cognitive scaffolding rather than cognitive substitution, particularly for individuals whose modes of thinking diverge from dominant norms, contributing to a more democratic epistemic practice.