valorisons

Valorisons are a fictional construct used in discussions of value theory and artificial intelligence ethics to quantify the alignment of actions, policies, or agents with a specified set of normative criteria. The term denotes either a singular valorison score or a multidimensional valorison vector that aggregates judgments about moral and practical worth.

A valorison typically encompasses dimensions such as harm minimization, fairness, autonomy respect, dignity, welfare, and sustainability. The score can be scalar or a vector, with context-dependent weighting to reflect priorities in a given scenario.

Measurement and computation are described in speculative literature as combining expert judgments, empirical data, and model predictions. Observations are mapped into standardized scales, and a normative function aggregates dimensions to produce the valorison. In AI alignment discussions, valorisons are proposed as a tool for steering decisions toward ethically bounded outcomes while maintaining overall utility.

The concept originated in contemporary analytic philosophy and speculative fiction as a way to examine how different normative theories might influence policy evaluation. It has appeared in thought experiments, design fiction, and discussions of algorithmic governance, though it remains theoretical and nonstandard in actual policy practice.

Critics argue that valorisons inherit cultural and subjective biases, risk simplification of complex ethics, and could be manipulated by framing or weighting choices. Proponents see them as a structured means to compare options and illuminate trade-offs in value-sensitive design.

See also value theory, normative ethics, AI alignment, decision theory, ethical metrics.