the next step is when the plant engineers cannot do what the AI can do even inefficiently. This becomes worse as expertise is reduced by dependency on AI. Then management becomes dependent on the AI. The root cause analyzer needs to be interpretable so that management can verify accuracy and goal achievement. Dependency on a black box process leads to management loss of control. Realistically even the best models are not fully reliable now even with hard data analysis. But I can see efficiency gains over time especially in preventive and predictive maintenance as reactive maintenance is reduced. Good podcast. Thanks.