Surveying Attitudinal Alignment Between Large Language Models Vs. Humans Towards 17 Sustainable Development Goals

2024-04-22

arXiv.org

AI Summary

This study identifies systematic misalignments between large language models (LLMs) and human attitudes across all 17 UN Sustainable Development Goals (SDGs), revealing critical gaps in conceptual understanding, affective valence imbalance, cultural-regional bias, and decision-logic mismatch—risks that may exacerbate social inequity, racial prejudice, and environmental harm. Method: We propose the first SDG-oriented LLM–human attitude alignment analytical framework, integrating cross-cultural semantic analysis, bias auditing, risk root-cause tracing, and an SDG knowledge graph to establish a multidimensional attitude annotation schema. Contribution/Results: Empirical analysis uncovers significant attenuation of support or value drift in LLMs for key SDGs—including Climate Action (SDG 13), Gender Equality (SDG 5), and Zero Hunger (SDG 2). Based on these findings, we formulate a data–training–evaluation co-optimization pathway and actionable alignment governance strategies, offering an SDG-grounded methodological foundation for ethical LLM alignment.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Humans and AI: Learning Human Values and Preferences

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

Large Language Models (LLMs) have emerged as potent tools for advancing the United Nations' Sustainable Development Goals (SDGs). However, the attitudinal disparities between LLMs and humans towards these goals can pose significant challenges. This study conducts a comprehensive review and analysis of the existing literature on the attitudes of LLMs towards the 17 SDGs, emphasizing the comparison between their attitudes and support for each goal and those of humans. We examine the potential disparities, primarily focusing on aspects such as understanding and emotions, cultural and regional differences, task objective variations, and factors considered in the decision-making process. These disparities arise from the underrepresentation and imbalance in LLM training data, historical biases, quality issues, lack of contextual understanding, and skewed ethical values reflected. The study also investigates the risks and harms that may arise from neglecting the attitudes of LLMs towards the SDGs, including the exacerbation of social inequalities, racial discrimination, environmental destruction, and resource wastage. To address these challenges, we propose strategies and recommendations to guide and regulate the application of LLMs, ensuring their alignment with the principles and goals of the SDGs, and therefore creating a more just, inclusive, and sustainable future.

Problem

• Large Language Models
• Human Differences
• Social Inequity

Innovation

• Large Language Models
• Sustainable Development Goals
• Bias and Fairness