人工智能与事实认定

人工智能与司法的深度融合,既体现在案件的法律适用环节也体现在事实认定环节,而事实认定是法律适用的前提。人工智能对案件事实认定的介入,需要将证据数据化、对数据进行运算整合、输出人可以理解的结论。在证据数据化环节,需要对证据进行结构化的数据改造,并克服语言障碍。在数据整合环节,人工智能主要以概率推理而不是因果推理作为逻辑推理方式,其算法也需要面对可计算性与复杂性两大难题。在结论输出环节,需要解决机器学习如何深化、信念如何建立与机器如何表达等难题。人工智能融入案件事实认定所面临的这些主要难题,可以尝试通过“小数据”训练,逐步构建人工智能“心智微结构”去慢慢攻克。

Artificial Intelligence and Fact-finding

The deep integration of artificial intelligence and justice is embodied in both the application of law and fact-finding.The latter is a precondition of the former.To intervene in the fact-finding in a case,artificial intelligence needs to digitize the evidence,integrate the data,and output conclusions that can be understood by human beings.On the step of digitalization of evidence,it is necessary to carry out structured data transformation of evidence and overcome language problems.On the step of data integration,artificial intelligence mainly uses probabilistic reasoning,rather than causal reasoning,as its logical reasoning model and needs to face two major problems of computability and complexity in its algorithm.On the step of conclusion output,it needs to solve the problems of how to deepen machine learning,how to build belief,and how the machine expresses itself.The main problems faced by the integration of artificial intelligence into the fact-finding in cases can be solved through“small data”training and the gradual building of“mind-microstructure”of artificial intelligence.