---
type: claim
source: "2025-12-01-fullstack-alignment-thick-models-value.md"
confidence: experimental
description: Thick models of value distinguish enduring values from temporary preferences.
created: 2025-12-01
processed_date: 2025-12-01
---

# Thick models of value distinguish enduring values from temporary preferences, enabling normative reasoning

The claim is based on a single paper that argues for the importance of distinguishing between enduring values and temporary preferences in AI alignment. This distinction is crucial for enabling normative reasoning within AI systems.

## Relevant Notes
- [[ai-alignment-requires-institutional-co-alignment-not-just-model-alignment]]