https://aifaultline.com/authors/joshua-s-gans/New papers by Joshua S. Gans in the AI Fault Line guide, and news naming them. https://aifaultline.com/research/p/title-0209d7613d9307e4/https://aifaultline.com/research/p/title-0209d7613d9307e4/This NBER paper models firms asking workers to train the AI systems that will replace them, where a worker can shirk, cooperate with training, or obstruct it while staying productive. Because finishing the training ends the job, the threat of dismissal loses force as a penalty, which shapes how fast automation can proceed. https://aifaultline.com/research/p/title-d56211e7a9037887/https://aifaultline.com/research/p/title-d56211e7a9037887/ https://aifaultline.com/research/p/title-c3c3f6ce69bbb802/https://aifaultline.com/research/p/title-c3c3f6ce69bbb802/Gans builds a contracting model of "O-ring" jobs where workers intrinsically enjoy certain tasks and voluntarily contribute unpaid hours, comparing complete task-bundle pricing against paid task floors that let workers top up their own time. The framework shows how automation interacts with task enjoyment in determining which contract form minimizes the wage bill and which tasks firms should automate. It matters for the AI debate because it formalizes a neglected margin: when AI automates tasks workers actually like, firms lose a source of free effort, so the labor-market effects of automation depend on task-level preferences, not just marginal products. https://aifaultline.com/research/p/title-b921e97feafcc9b2/https://aifaultline.com/research/p/title-b921e97feafcc9b2/Gans (NBER) argues that standard estimates of AI's wage effects err by holding workers' task bundles fixed at their pre-AI shape: when automation breaks a job apart, firms endogenously choose how to rebundle the surviving tasks, and that rebundling margin drives much of the wage outcome. This theoretical point cuts at the foundation of the widely used exposure-index literature, suggesting measured wage effects conflate automation with firms' organizational responses and that job redesign, not task displacement alone, determines who gains or loses from AI.