Hacker News new | ask | show | jobs
by rshanreddy 193 days ago
This is a 1250 word judging prompt - likely AI generated Along with 10 scored conversation samples - all also AI generated No verification in the field, no real data

In other words, AI scores on AI conversations - disguised as a means of gauging clinical competence / quality?

This is not an eval - this is a one-shotted product spec!