We provide a framework for categorizing the ways in which conflicting incentives might lead LLMs to change the way they interact with users, inspired by literature from linguistics and advertising regulation
这项研究的创新之处在于将语言学和广告监管领域的理论应用于AI利益冲突分析,为理解和解决AI商业化中的伦理问题提供了新的理论框架,具有跨学科的重要意义。