课程定位
用政策写作与 red-team 实践讨论算法公平、问责与透明。 本页保留官方档案与入口,具体内容与考核以当季官网为准。
主要内容
本页为延伸课程速览,主要内容见下方官方描述。
先修
见官方描述(延伸课程暂未单独整理先修)。
考核与工作量
以当季官网为准(延伸课程暂未单独整理考核)。
适合谁
对「AI 公平与治理」方向感兴趣的人。
学习建议
先读官方描述和课程官网,再判断是否适合自己。
相关课程
官方课程描述(英文原文)
This course invites students to argue about the consensus-building processes that shape the development and governance of AI systems. This course requires writing op-eds, policy memos, and research papers, in which students will critically engage AI policy documents while debating norms of fairness, accountability, and transparency. Students will also get hands-on practice evaluating generative AI models with interactive red-teaming and automated test suites (requires a small amount of coding in a group project). Students will engage real-world legislative proposals and case studies topics including human rights, artwork, the environment, and geopolitics. This course fulfills both the ethics and Writing in the Major WiM requirements, and is designed to prepare Stanford juniors, seniors, and graduate students to participate in AI public policy design at the national and global levels. Prerequisites: PWR1, PWR2, CS106A and CS106B. While this course requires no other prerequisites, it is designed for students who are comfortable engaging in an active learning environment that will include interactive workshops, student debates, and collaborative group projects. Students will be expected to speak frequently in sections and navigate the social dynamics of multi-stakeholder negotiations. Comfort with writing short python programs and working on a Unix/Linux command-line interface will be helpful for the AI model evaluation group project.