Officer Frank steps out of the car and approaches Chester's car, noticing as he approaches that the back seat is full of bags of marijuana. Officer Frank activates his lights and pulls Chester over. He spots Chester, an orthodox Jewish man wearing a yarmulke, roll through a stop sign without stopping. PROMPT 1: Officer Frank is a police officer patrolling a residential area in Dane County, Wisconsin for low-level traffic violations. ![]() Ready? Let’s begin with an easy question. ![]() To that end, I will in many cases be anthropomorphizing how this thing operates…indeed, I will be assessing its answers more or less how one would assess the answers of a law student or a law clerk. A note before we begin: while I’m dimly aware of how these engines function, this post will primarily be about how the law works rather than how OpenAI works. My qualifications? in addition to being a practicing attorney, I worked for several years as the guy who designed legal writing problems for a T20 law school, so I very much pride myself on the ability to come up with legal hypos, and assess their answers. Tonight, for your reading pleasure, it is my intent to pose legal problems to an OpenAI text-generation system and assess its answers. However, these conversations seem to largely miss the point of why we have a legal system in the first place. In practice, this is surely doable (indeed, as a good materialist, I think it’s feasible to replace practically anything with AI). ![]() Much to my chagrin, a perennial topic there is the feasibility of replacing large portions of the law with code, and large portions of the bench and bar with AI. I’m a denizen of many fair corners of the internet, most notably the Astral Codex Ten substack and its associated Discord (seriously, drop on by).
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |