This study explores the potential and limitations of generative artificial intelligence, with a particular focus on ChatGPT-4, in reproducing human values and attitudes based on socio-demographic profiles. By comparing real data from the Euro-pean Social Survey (ESS) with AI-generated data, the study assesses the ability of AI to reflect public opinion trends. While AI-generated responses exhibit a general alignment with survey data, they show limited variability and some inconsistencies in group-level analyses. In conclusion, despite its promising aspects, the findings suggest that AI cannot replace empirical research.

Artificial social research? AI's capabilities and risks in predicting values and attitudes

Domenico Trezza
2025-01-01

Abstract

This study explores the potential and limitations of generative artificial intelligence, with a particular focus on ChatGPT-4, in reproducing human values and attitudes based on socio-demographic profiles. By comparing real data from the Euro-pean Social Survey (ESS) with AI-generated data, the study assesses the ability of AI to reflect public opinion trends. While AI-generated responses exhibit a general alignment with survey data, they show limited variability and some inconsistencies in group-level analyses. In conclusion, despite its promising aspects, the findings suggest that AI cannot replace empirical research.
2025
artificial intelligence, experiment, predictive capacity, validity and reliability
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12607/73786
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
social impact