Thick Alignment: A Social Framework for Values and AI Systems

arXiv:2611.05857v1 Announce Type: new Abstract: Computer science research on the "alignment" of artificial intelligence (AI) systems treats human values as content to be specified and optimized. Decades of social science scholarship describe values differently, as enacted in ordinary practice, interpreted in social context, and contested across plural orders of evaluation and worth. We argue that the dominant alignment techniques, which we call "thin alignment," convert situated moral or value judgments into one or a few metrics via devices of commensuration such as preference labeling, reward modeling, and consumer ratings. Drawing on philosophical and anthropological accounts of thickness, "thick alignment" instead examines how values take shape, inform the development of AI systems, and become subject to contestation over whose interests they serve. We describe two forms of misalignment in seemingly aligned algorithmic systems that highlight thin alignment: "Halcyon misalignment," which presumes a prior, just equilibrium of technology and values from which a system has departed, and "Mirrored misalignment," which names the risks and harms produced by systems that are aligned in a thin sense yet accurately reproduce an unequal social order. A thick account registers such systems as misaligned. We then trace how researchers have taken up and adapted the concept of thick alignment since it was first proposed in 2023, using these adaptations to refine the concept, and to distill three of its pillars: (1) contextual description that both reflects and constitutes values; (2) i

Critique of current AI alignment techniques and introduction of 'thick alignment' concept highlights potential risks and harms of AI systems.

The paper critiques current AI alignment methods for oversimplifying human values and proposes a richer framework ("thick alignment") that includes social context and contestation. This approach could reduce harms from misaligned AI reproducing social inequalities but also reveals complex challenges for aligning AI with diverse, contested values. The score reflects cautious optimism that this framework could improve AI ethics and governance, though practical impact is uncertain.

This is a theoretical and conceptual contribution critiquing alignment methodology, not an empirical measurement of societal harm or benefit. It raises legitimate concerns about alignment reproducing inequality but offers no concrete evidence of deployed impact, so the implication for flourishing is modest and indirect.

This paper outlines a conceptual framework identifying significant flaws in current AI alignment approaches. It claims existing methods can perpetuate or even exacerbate social inequalities, which represents a concerning implication for human flourishing. The paper is academic and theoretical, proposing a new way to understand existing problems rather than presenting new empirical findings.

The paper proposes a framework for aligning AI systems with human values in a more nuanced and socially aware way. It critiques current methods and offers an alternative that could lead to more equitable AI systems. The implications are positive for human flourishing if adopted.

SOURCE     arXiv cs.CY (Computers and Society) (tier 1)
   ↓
DOCUMENT   238b3b3d-50b1-4348-a2f2-136c8177ce99
           https://arxiv.org/abs/2611.05857
   ↓
EVIDENCE   2 extracted excerpts
   ↓
MODEL RUN  5 runs, methodology 1.0.1
   ↓
SCORE      0  (Mixed or uncertain)
   ↓
CONFIDENCE 59%