Simulating the End of Work and Money: A Microsimulation of AGI-Driven Automation

The potential for Artificial General Intelligence (AGI) to automate a wide range of economically relevant tasks raises important questions about the future of labor markets. This paper presents a microsimulation framework in which firms iteratively adjust their automation levels within a repeated non-cooperative game, structurally analogous to the Prisoner's Dilemma. Firms balance the competitive advantages of out-automating rivals against the costs of implementation, leading to context-dependent equilibrium outcomes. Simulations indicate that automation adoption patterns vary with competitive pressure and cost structures: high competitive intensity is associated with near-complete automation in some sectors, while higher implementation costs are associated with partial labor substitution in others. These dynamics suggest that unregulated competitive interactions may generate socially suboptimal automation levels, motivating the exploration of policy instruments such as automation taxation or universal basic income mechanisms funded by AGI-generated rents. The study contributes to the literature by isolating competitive incentives as a potential driver of automation, distinct from purely technological or macroeconomic determinants, and by highlighting the role of institutional design in shaping the economic impact of AGI.