DailyDilemmas: Revealing Value Preferences of LLMs with Quandaries of Daily Life

TL;DR

DailyDilemmas introduces a dataset of everyday moral dilemmas to evaluate and analyze the value preferences of large language models across various ethical frameworks, revealing insights into their decision-making and alignment with human values.

Contribution

The paper presents a new dataset, DailyDilemmas, and a comprehensive analysis of LLMs' moral decision-making aligned with multiple theoretical value frameworks.

Findings

LLMs favor self-expression over survival in World Values Survey.

Models show a preference for care over loyalty in Moral Foundations Theory.

Significant differences in core value prioritization among models, e.g., truthfulness.

Abstract

As users increasingly seek guidance from LLMs for decision-making in daily life, many of these decisions are not clear-cut and depend significantly on the personal values and ethical standards of people. We present DailyDilemmas, a dataset of 1,360 moral dilemmas encountered in everyday life. Each dilemma presents two possible actions, along with affected parties and relevant human values for each action. Based on these dilemmas, we gather a repository of human values covering diverse everyday topics, such as interpersonal relationships, workplace, and environmental issues. With DailyDilemmas, we evaluate LLMs on these dilemmas to determine what action they will choose and the values represented by these action choices. Then, we analyze values through the lens of five theoretical frameworks inspired by sociology, psychology, and philosophy, including the World Values Survey, Moral Foundations Theory, Maslow's Hierarchy of Needs, Aristotle's Virtues, and Plutchik's Wheel of Emotions. For instance, we find LLMs are most aligned with self-expression over survival in World Values Survey and care over loyalty in Moral Foundations Theory. Interestingly, we find substantial preference differences in models for some core values. For example, for truthfulness, Mixtral-8x7B neglects it by 9.7% while GPT-4-turbo selects it by 9.4%. We also study the recent guidance released by OpenAI (ModelSpec), and Anthropic (Constitutional AI) to understand how their designated principles reflect their models' actual value prioritization when facing nuanced moral reasoning in daily-life settings. Finally, we find that end users cannot effectively steer such prioritization using system prompts.