Law Professors Prefer AI Over Peer Answers
| hilarious cyan hell | 06/02/26 | | white idiot | 06/02/26 | | Adventurous ticket booth | 06/02/26 | | hilarious cyan hell | 06/02/26 | | disrespectful senate brethren | 06/02/26 | | Snowy round eye | 06/02/26 | | hilarious cyan hell | 06/02/26 | | disrespectful senate brethren | 06/02/26 | | dark mediation | 06/03/26 | | floppy big-titted voyeur stock car | 06/02/26 | | Adventurous ticket booth | 06/02/26 | | dark mediation | 06/03/26 | | cruel-hearted den trust fund | 06/02/26 |
Poast new message in this thread
Date: June 2nd, 2026 5:12 PM Author: hilarious cyan hell
https://law.stanford.edu/wp-content/uploads/2026/06/salinas_et_al.pdf
"Professors rated LLMs far higher than their peers (average win rate = 75.33%), with models performing similarly to the best instructor. LLM responses were also rarely flagged as harmful (3.53%, vs 12.06% for professors). Preferences for LLM answers were consistent across evaluators and reflected shared professional standards."
(http://www.autoadmit.com/thread.php?thread_id=5870655&forum_id=2)#49911519) |
Date: June 2nd, 2026 8:07 PM Author: cruel-hearted den trust fund
This is a dumb study with a predictable result.
They're essentially taking popular 1L cases taught first year law students, and asking professors and LLMs to explain them.
Predictably, the LLMS did better because they can just summarize the casebooks from the past 40 years in the most concise way.
But let's say we a ran a study, where we compared the explanations in the study guide vs the explanations of professors, and other professors chose the study guide 75% of the time. Would that be meaningful or surprising?
I wouldn't think so...
(http://www.autoadmit.com/thread.php?thread_id=5870655&forum_id=2)#49911897) |
|
|