Language-based game theory in the age of artificial intelligenceCapraro, Valerio (Avtor) Di Paolo, Roberto (Avtor) Perc, Matjaž (Avtor) Pizziol, Veronica (Avtor) game theoryartificial intelligencesocial preferenceslanguage-based preferencessocial physicsmoral behaviortrustUnderstanding human behaviour in decision problems and strategicinteractions has wide-ranging applications in economics, psychology andartificial intelligence. Game theory offers a robust foundation for this under-standing, based on the idea that individuals aim to maximize a utilityfunction. However, the exact factors influencing strategy choices remainelusive. While traditional models try to explain human behaviour as a func-tion of the outcomes of available actions, recent experimental researchreveals that linguistic content significantly impacts decision-making, thusprompting a paradigm shift from outcome-based to language-based utilityfunctions. This shift is more urgent than ever, given the advancement ofgenerative AI, which has the potential to support humans in making criticaldecisions through language-based interactions. We propose sentiment analy-sis as a fundamental tool for this shift and take an initial step by analysing61 experimental instructions from the dictator game, an economic gamecapturing the balance between self-interest and the interest of others,which is at the core of many social interactions. Our meta-analysis showsthat sentiment analysis can explain human behaviour beyond economicoutcomes. We discuss future research directions. We hope this worksets the stage for a novel game-theoretical approach that emphasizes theimportance of language in human decisions.Royalsociety publishing20242024-11-27 13:04:26Znanstveno delo91197sl