Computational Ethics and Large Language Models

4 units · Graduate courses (200-299)

Lecture, four hours; outside study, eight hours. Preparation: a basic understanding of artificial intelligence and natural language processing concepts is encouraged. Large language models (LLMs) are becoming ubiquitous. They are used in many real-world applications, ranging from content moderation and online advertisement to health care. Given their increasing role in what people see, how they think, and what is publicly known about them, it is critical to consider the ethical ramifications of deploying LLM-based systems. Study offers a lens on historical and current computational ethics problems in natural language processing (NLP). Reading and discussion of literature about how large-scale language modeling is impacting domains such as privacy, health care, and political science. Discussions are accompanied by guest lectures from domain experts. In group projects students develop critical thinking and problem-solving through hands-on data curation and writing short perspective pieces on a future AI policy framework to mitigate ethical risks observed from the literature review. Letter grading.