The paper argues that policy makers should not use the term artificial intelligence (AI) to define the material scope of AI regulations. The argument is developed by proposing a number of requirements for legal definitions, surveying existing AI definitions, and then discussing the extent to which they meet the proposed requirements. It is shown that existing definitions of AI do not meet the most important requirements for legal definitions. Next, the paper suggests that policy makers should instead deploy a risk-based definition of AI. Rather than using the term AI, they should focus on the specific risks they want to reduce. It is shown that the requirements for legal definitions can be better met by considering the main causes of relevant risks: certain technical approaches (e.g. reinforcement learning), applications (e.g. facial recognition), and capabilities (e.g. the ability to physically interact with the environment). Finally, the paper discusses the extent to which this approach can also be applied to more advanced AI systems.
翻译:文件认为,决策者不应使用人工智能(AI)这一术语来界定AI规章的物质范围,而应使用人工智能(AI)这一术语来界定AI规章的物质范围;提出这一论点的方式是提出法律定义的若干要求,调查现有的AI定义,然后讨论这些定义满足拟议要求的程度;文件表明,AI的现有定义不符合法律定义的最重要要求;接着,文件建议决策者应采用AI的基于风险的定义;而不是使用AI这一术语,它们应侧重于他们希望减少的具体风险;文件表明,通过考虑相关风险的主要原因,可以更好地满足法律定义的要求:某些技术方法(例如加强学习)、应用(例如面部识别)和能力(例如与环境进行身体互动的能力);最后,文件讨论了这一方法在多大程度上也可适用于较先进的AI系统。