人工智能时代的法律议论

Legal Argumentation in the Era of Artificial Intelligence

人工智能作为一种重要的技术手段,能够辅助法律议论,确保法律论证、推理、判断以及决定的客观性和中立性。但是,司法人工智能在提高同案同判水平和审判可预测性的同时,也容易导致法律议论流于形式,助长算法歧视。人工智能的深度学习离不开法律专家的介入和监督,智慧司法的系统构建应为法律解释和法律议论预留空间。为使人工智能真正有效地模拟法律议论,法律人需对法律背景知识体系进行梳理和电脑化处理,从一般条款、元规则入手,以法律论题学为媒介,建立价值标准体系。为实现价值判断的客观化,除为法律推理的价值体系建立论题目录、缔结关于法律议论的通信协议外,确立价值函数和价值权重也是一项重要任务。此外,还应开拓与人工智能相兼容的法律议论方式,将具体的场景和语境纳入法律议论电脑化的视野,开发相应的技术方法和模型。

As an important technical means, AI can assist legal argumentation and ensure the objectivity and neutrality of legal argumentation, reasoning, judgment and decision. In helping to improve the accuracy of "treating like cases alike", as well as the predictability of the trial process, judicial AI may in a similar way lead to the result of legal argumentation being hollowed out and becoming no more than mere formality, as well as fostering algorithmic discrimination. For this very reason, the deep learning of AI is inseparable from the intervention and supervision of legal experts. The system construction of intelligent justice should reserve space for legal interpretation and legal discussion. To make artificial intelligence simulate the real law effectively, lawyers should sort out legal background knowledge and enable the computerized process by virtue of general rules and meta rules, by means of the catalogue of topics, and finally establish a value standard system. In order to achieve the objective of value judgment, in addition to build the value system of legal reasoning topic directory to conclude about the law of communication protocols, establishing the value function and the value of the weight is also an important task. Furthermore, there is a need to develop the methods of legal argumentation that are compatible with AI technique, bring the specific scenarios and context into the visual field of the computerization of legal argumentation, and develop corresponding technical methods and models.