Institutional Repository of Vilnius University
Dirbtinio intelekto konsultavimo naudojant chatgpt įtaka žaidėjų strategijoms ir rezultatams ultimatumo žaidime: lyginamoji analizė /
Abstract
dc:descriptionThere has been an increase in the growth of use of Artificial Intelligence (AI) models in the past few years. AI models strike similarities with human beings in the way they reason, communicate and make decisions. The field of behavioral economics can vastly benefit by having AI as a determining factor to how humans base their decisions in strategic contexts such as the Ultimatum Game, a fundamental experiment in economic behavior. This thesis investigates the impact of AI advisement to human decision-making in a continuous Ultimatum Game. This thesis attempts to examine the variations in offer amounts, acceptance rates and advisement rates in varying rounds of the continuous Ultimatum Game. I attempted to assess the impact of AI advisement on players’ decision-making strategies in a qualitative manner. In addition, I attempted to analyze the outcomes in a more numerical analysis. The thesis utilizes a combination of qualitative approaches to the findings and more mathematically sound tests, such as the ANOVA and Chi-Squared tests to analyze the data collected. Participants played three rounds of the Ultimatum Game, with varying conditions, in which AI advisement was included. I found that there was a 17.36% increase in mean offers from round 1 to 2, in which the proposer had the option to advise with the AI model ChatGPT. However, ANOVA and Chi-Squared tests showed that the results were statistically insignificant. Nonetheless, the increase is visible and apparent, indicating that AI advisement promoted players to be more fair and be more strategic in their decisions.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Bučinskas, Julius,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210640384&prefLang=en_US