LESSWRONG
LW

Wikitags

Preference framework

Edited by Eliezer Yudkowsky, et al. last updated 14th Feb 2017

A 'preference framework' refers to a fixed algorithm that updates, or potentially changes in other ways, to determine what the agent for outcomes. 'Preference framework' is a term more general than '' which includes structurally complicated generalizations of utility functions.

As a central example, the proposal has the agent switching between utility functions UX and UY depending on whether a switch is pressed. We can call this meta-system a 'preference framework' to avoid presuming in advance that it embodies a utility function.

An even more general term would be which doesn't presume that the agent operates by .

Parents:
Children:
and 1 more
prefers
2
2
utility function
Discussion0
Discussion0
utility indifference
VNM-coherent
terminal
Attainable optimum
Decision_algorithm
Value alignment problem
Moral uncertainty
preferring outcomes