angenoev 2 karma 23d on HN HN profile →
Epistemic risk measurement in AI systems. Building ONTO Standard — open measurement protocol for AI structural discipline.
Coverage
We've seen 1 of ~2 submissions
Full eval: 0 Lite-only: 0 Unevaluated: 1
1 stories
1. We gave AI models epistemic discipline. They started describing themselves (medium.com)
3 points by angenoev 8 days ago | 1 comments | skipped