A System Dynamics and NSGA-II Based Framework for Occupational Structure Evolution in the Game Industry under Generative AI

Abstract

To characterize the impact of generative AI technology development on changes in job task structures and requirements, this paper constructs a multi-level algorithmic analysis framework. First, a three-dimensional feature vector space is constructed based on O*NET data, and representative job types are selected using the feature distance maximization method. Subsequently, a multidimensional evaluation model for task decomposition is proposed, incorporating indicators such as routine nature, cognitive complexity, physical dependency, and creativity into an evaluation vector. Weights are determined using the Analytic Hierarchy Process (AHP), and a nonlinear substitution rate model incorporating a Sigmoid threshold function is constructed to calculate the job impact index. Building upon this foundation, a Logistic growth function characterizes technological maturity. Combined with system dynamics, this establishes an evolutionary model for job demand, numerically solved via the fourth-order Runge–Kutta method to simulate medium-to-long-term demand changes. Concurrently, grey relational analysis identifies key drivers, and a multi-objective optimization model under Stackelberg games is constructed. The NSGA-II algorithm is employed to obtain Pareto optimal solutions. The results demonstrate that this methodology can characterize the synergistic effects of technological substitution and efficiency gains within a unified computational framework, providing a general modeling approach for analyzing job structure evolution in complex technological environments.