Although artificial intelligence (AI) technology has obvious advantages in improving employee performance, the mechanism of its efficiency improvement has rarely been systematically explored. This study aims to investigate the mechanism by which AI technology affects the work efficiency of manufacturing employees. Combining technological determinism and the job characteristics model, it selects job characteristics such as knowledge specialization, job demands, job autonomy, and social support from the perspectives of knowledge, tasks, and society as mediating variables, and intrinsic motivation as a moderating variable, to construct a technology impact model. A three-step mixed method is used for path analysis and neural network model verification. After controlling for substitution crisis and demographic variables, the study finds that the application of AI can enhance employee work efficiency through job demands, job autonomy, or knowledge specialization respectively; the influence effect of job demands exceeds that of the other two variables; social support cannot play a mediating role in the relationship between AI application and employee work efficiency, but intrinsic motivation has a negative moderating effect on such relationship. This conclusion helps to understand the implementation mechanism of AI technology in the manufacturing industry and provides a path reference for continuously improving the efficiency of human-machine collaboration.