AI Use Divergence and Efficiency Norm Spillover: Intergenerational Accommodation, Status Judgments, and Autonomy Conflict as Technology Becomes Cognitive Infrastructure

As generative AI becomes part of everyday information seeking and decision support, people diverge in access, usage intensity, task competence, and beliefs about model reliability. This conceptual paper examines how such differences may become status signals and expectations about how others ought to think. It proposes Efficiency Norm Spillover, a testable process in which context-specific productivity gains are generalized into demands for others to adopt a tool. A reflexive single-case analysis of an intergenerational disagreement is situated alongside industrial time discipline, computerization, older adults' digital exclusion, and evidence on AI productivity and social evaluation. The paper further proposes Bidirectional Cognitive Accommodation as a research hypothesis: experienced users adjust communication and support; less experienced users may learn tasks of personal value without coercion; and platforms and institutions ensure accessible design and non-AI alternatives. The case documents adoption demands, relationship-conditional language, conflict, and later reflection, but does not establish causality. Technical adoption is distinct from general intelligence or social worth. Falsifiable propositions and ethically constrained designs are offered to investigate when personal workflow preferences become social obligations and how the resulting barriers can be reduced.