There is some background theory S the agent reasons with, say ZFC. This theory is extended by definitions to define action A and utility U. Say, these extensions consist of sets of axioms AX and UX. Then, the agent derives the set of moral arguments M from theory S+AX+UX. By preference, I refer specifically to UX, which defines utility U in the context of agent’s theory S. But if M is all (moral arguments) the agent will infer, then S+AX+M also defines U, just as well as S+AX+UX did. Thus, at that point, we can forget about UX and use M instead.
I’m not sure why you want to think in terms of S+AX+M instead of S+AX+UX, though. Doesn’t starting with the axiom set S union AX union UX better reflect how the agent actually reasons?
It does start with S+AX+UX, but it ends with essentially S+AX+M. This allows to understand the point of this activity better: by changing original axioms to equivalent ones, the agent expresses the initially separately defined outcome in terms of action, and uses that expression (dependence) to determine the outcome it prefers.
There is some background theory S the agent reasons with, say ZFC. This theory is extended by definitions to define action A and utility U. Say, these extensions consist of sets of axioms AX and UX. Then, the agent derives the set of moral arguments M from theory S+AX+UX. By preference, I refer specifically to UX, which defines utility U in the context of agent’s theory S. But if M is all (moral arguments) the agent will infer, then S+AX+M also defines U, just as well as S+AX+UX did. Thus, at that point, we can forget about UX and use M instead.
Okay, thanks. This is clear.
I’m not sure why you want to think in terms of S+AX+M instead of S+AX+UX, though. Doesn’t starting with the axiom set S union AX union UX better reflect how the agent actually reasons?
It does start with S+AX+UX, but it ends with essentially S+AX+M. This allows to understand the point of this activity better: by changing original axioms to equivalent ones, the agent expresses the initially separately defined outcome in terms of action, and uses that expression (dependence) to determine the outcome it prefers.