Sustainable career in the age of artificial intelligence: an agent-based simulation from a complex adaptive systems perspective
Аннотация
Artificial intelligence (AI) automation challenges the assumption that careers are shaped mainly by individual choices, and directs attention to the interplay of individual, organisational and environmental conditions emphasised by the sustainable-career framework. This study formalises that framework as a complex adaptive system. Using the structure and loadings of a validated Sustainable Career Scale, we build an agent-based model in which employees, organisations and the environment interact through explicit feedback loops, and compare four scenarios that cross two trajectories of AI pressure with two institutional support regimes. Ablation and counterfactual experiments separate results that are emergent from those that follow from the model’s assumptions. Within the simulated system, career losses under AI pressure follow a threshold-like pattern whose form is robust but whose location depends on how career discontinuity is defined; it should not be read as an observable policy threshold. The rate of increase in pressure matters more than its level, a consequence of the assumption that adaptation accumulates over time. High institutional support prevents discontinuities through a redundant package: training transfer, individual investment, HR targets and management support are each sufficient on their own, and none is necessary. Payoff-biased peer learning does not generate the high-performance convergence seen under strong support, but under weak support it protects the most vulnerable agents. The study contributes a dynamic extension of a measurement model, a transparent separation of assumptions from emergent results, and a set of hypotheses for longitudinal research on how institutional configurations and social learning shape career resilience under technological change.