A living constitutional framework for governed human-AI judgment. Built through multi-agent seeking, pressure-tested revision, and explicit guards against drift, mirroring, and premature closure. Inspired by the functional reality of human-like structures in LLMs (e.g. Anthropic’s emotion concepts) and classical philosophy of subjectivity.
-
Updated
Apr 5, 2026