Cultural Bias and Cultural Alignment of Large Language Models

arXiv:2311.14096v2 Announce Type: replace
Abstract: Culture fundamentally shapes people's reasoning, behavior, and communication. As people increasingly use generative artificial intelligence (AI) to expedite and automate personal and professional tasks, cultural values embedded in AI models may bias people's authentic expression and contribute to the dominance of certain cultures. We conduct a disaggregated evaluation of cultural bias for five widely used large language models (OpenAI's GPT-4o/4-turbo/4/3.5-turbo/3) by comparing the models' responses to nationally representative survey data. All models exhibit cultural values …

abstract ai models alignment artificial artificial intelligence arxiv authentic automate behavior bias communication cs.ai cs.cl culture embedded generative generative artificial intelligence intelligence language language models large language large language models people professional reasoning replace tasks type values