Taught from a litigator's perspective, this course examines using LLMs in everyday practice. The course will teach how attention, the skill that defines great advocacy, bridges legal practice and the transformer architecture powering LLMs. Lawyers already structure arguments, visuals, and case narratives to direct attention; large language models allocate it through a structurally parallel process. A model's context window demands the same discipline as a jury's attention span: structure, selection, and sequence matter. From that foundation, students develop a practitioner's understanding of transformers, tool use, reasoning, and agentic systems, and apply them to case development, context assembly, strategic analysis, and advocacy. Students then design and build their own AI-powered workflows for case development and analysis. The course shows why the litigator's craft is already the foundation of AI-augmented practice, and gives students the tools to prove it. What remains irreducibly human: judgment, strategy, and framing. No coding required. Elements used in grading: Paper.
1 units · Law Honors/Pass/Restrd Cr/Fail
Taught from a litigator's perspective, this course examines using LLMs in everyday practice. The course will teach how attention, the skill that defines great advocacy, bridges legal practice and the transformer architecture powering LLMs. Lawyers already structure arguments, visuals, and case narratives to direct attention; large language models allocate it through a structurally parallel process. A model's context window demands the same discipline as a jury's attention span: structure, selection, and sequence matter. From that foundation, students develop a practitioner's understanding of transformers, tool use, reasoning, and agentic systems, and apply them to case development, context assembly, strategic analysis, and advocacy. Students then design and build their own AI-powered workflows for case development and analysis. The course shows why the litigator's craft is already the foundation of AI-augmented practice, and gives students the tools to prove it. What remains irreducibly human: judgment, strategy, and framing. No coding required. Elements used in grading: Paper.
Offered in Autumn 2026 at Stanford University.