It is a question worthy of serious consideration that whether AI, which is an economic growth point driven by technology, would become the growth point of the rule of law. The essence of AI is a set of programs, which are composed of algorithms and data structures. It does not lead to the emergence of good law, because good law is a social practice problem and is inseparable from value judgment. As a set of human material conditions, the algorithm lacks the ability to practice both at philosophical and actual level. AI reflects strategic thinking that emphasizes predictive control of procedure, but the trial is a normative thinking that emphasizes the expressive function of the procedure. The judges rely on their faith of the law rather than their memories to judge the cases. This non-gaming behavior cannot be followed by AI. AI can store and analyze massive amounts of data, but what the justice needs is experience. The experience is the internalization of human life rather the coded arrangement of data. Both are incommensurable and irreplaceable. Needless to say, AI has played an active role in judicial reform, but in order to further deepen reforms, we must also see its shortcomings. The realization of the rule of law should pay more attention to system construction, and AI should exert its own advantages on this basis.