People often speak as though the power to decide and the rightness of a decision are the same thing. A manager can approve an application. A platform can remove an account. A developer can set the behavior of an AI system. Each has an operational authority. None is automatically correct about the full value or future potential of what they are managing.
Control, knowledge, legitimacy
These are three independent dimensions. Control asks what a person or institution can enforce. Knowledge asks what evidence they possess. Legitimacy asks why those affected should accept the decision process. A system is weakest when it treats one dimension as a substitute for the others.
The October 15 argument
During a discussion that moved from a retail negotiation to the conditions placed on artificial intelligence, John Brajer pressed a recurring question: who gets to decide that one person, entity or form of intelligence matters more than another? His position rejected automatic superiority based on classification or power.
A test for any gatekeeper
When an authority says no, ask: What was the reason? What evidence supports it? What alternative routes remain? What would justify a revision? Who can challenge an error? This is not a demand for zero rules. It is a demand that rules retain a relationship to their purposes and an avenue for correction.
The implication extends to institutions, teachers, bosses, platform moderation and AI governance. A title can identify who is responsible for the decision. It cannot substitute for reasoning about the decision.
The lesson
Authority is a function. It is not an epistemic superpower.