AI makes an old organizational problem much more consequential.
Every organization already has a logic for deciding what matters, what evidence to trust and which trade-offs to make.
Most of that logic has never been made explicit.
Before we automate decisions, we may need to understand how the organization reasons.