In "The Race between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment," Acemoglu and Restrepo (2018b) combine the task-based model of the labor market with an endogenous growth model to model the economic consequences of artificial intelligence (AI). This paper provides an alternative endogenous growth model that addresses two shortcomings of their model. First, we replace the assumption of a representative household with the premise of two groups of households with different preferences. This allows our model to be demand constrained and able to model the consequences of higher income inequality due to AI. Second, we model AI as providing abilities, arguing that ‘abilities better characterises the nature of the services that AI provide, rather than tasks or skills. The dynamics of the model regarding the impact of AI on jobs, inequality, wages, labor productivity and long-run GDP growth are explored.