Simulating human well-being with large language models: Systematic validation and misestimation across 64,000 individuals from 64 countries

Proceedings of the National Academy of Sciences, Volume 122, Issue 48, December 2025.
SignificanceGovernments and international organizations increasingly view subjective well-being as a core measure of progress, yet high-quality data remain costly, uneven, and infrequent. Large language models (LLMs) have been proposed as scalable …
Read More