The Atlantic's piece on America heading toward the infinite workweek lands differently when read alongside two arXiv papers published the same day. Researchers have now built SkillChain-Gym, a benchmark explicitly designed for 'reskilling-aware production-inventory control,' and a companion paper on Skill-Constrained Model Predictive Control for manufacturing supply chains. Together they reveal something The Atlantic's cultural framing gestures at but does not name: the workforce is already being abstracted into a decision variable inside optimization models.
When Labor Becomes a Parameter
SkillChain-Gym treats 'qualified human capacity' as a formal constraint in production planning, something to be modeled, predicted, and optimized around disruptions. This is not sinister on its face. Supply chain resilience requires accounting for what workers can actually do. But the framing is revealing. When human skill is a 'decision variable,' the question of what it feels like to be reskilled on a model's schedule does not compute. The infinite workweek The Atlantic describes is the phenomenological experience of living inside someone else's optimization function.
Exhaustion as an Emergent Property
Fast Company's reporting on corporate America failing working mothers adds a dimension the academic models miss entirely: the optimization assumes an unconstrained worker. Mothers, caregivers, anyone with obligations outside the production function, are not edge cases. They are the majority. The AI agents The Atlantic envisions will not create exhaustion as a side effect. They will create it as an emergent property of a system that was never designed to model the full cost of human time. Fred Turner and Jasmine Sun's conversation on AGI as warmed-over network-society rhetoric is useful here: the infinite workweek is not a new problem made worse by AI. It is an old fantasy of total productivity, now given new tools.