Dirbtinio intelekto potencialas rinkimuose /

Voting has inherent flaws and problems that affect the validity and reliability of elections. Given the criticism of the current elections and the research that has been done in this field, we can understand that voting will never be perfect, and that it has many challenges, but a thoughtful reform based on the latest technological trends can help and improve our voting practices. One such possible example would be to replace voting with a sufficiently powerful artificial intelligence that makes a decision for us, based on the data it collects about us. On this basis, we formulate the main objective: people do not know their true preferences, only what information they declare themselves. In this way, elections become vulnerable and inaccurate, which undermines the integrity and credibility of elections, because they do not effectively reveal the preferences of voters. The current technology ecosystem with its artificial intelligence allows for a different voting experience, but are the latest technologies ready to change established voting norms? The main goal of this work is to investigate the ability of artificial intelligence to match human values to data and to rationally assign them to a party's stated values. In this way, selecting the party that is most in line with a person's internal preferences. To achieve this, I have set myself a number of key tasks: first, I provided an overview of the current electoral system, presenting the constraints and the main problems of the elections, together with the alternatives that already exist, such as convenience voting or lotocracy. I will then analyze the concept of artificial intelligence and its possible application in elections. To investigate the potential of AI, I will conduct a study that will try to clarify the limits, principles and reliability of AI, and I will include a voting advice machine that will allow a more accurate assessment of the potential of the technology in voting. Then I will analyze the results of the study and draw conclusions. Alongside this, I have set out the main working hypothesis: Artificial Intelligence - Chat GPT can use the data provided to identify a party that is consistent with a person's stated values. In order to achieve the highlighted goal, I carried out a study which was experimental and based on a thought experiment by E.B.Chapman. This study revealed important insights not only about artificial intelligence but also about VAA. The aforementioned research hypothesis did not prove to be true, as artificial intelligence cannot properly match the values declared by humans with the values declared by the parties according to the data provided. This opinion is based on the following findings: Artificial intelligence does not take into account the main cleavages in Lithuania, it evaluates the information it receives on a global level, it relies on classical dimensions without taking into account the historical context of Lithuania. The inaccuracy of AI is also due to misjudgement in foreign policy. In the context of the Russia- Ukraine war, Lithuanian citizens' opinions towards the Eastern neighbors were more severe, the majority of them expressed strong opposition towards Russia. In assessing the responses the artificial intelligence made inaccurate inferences about human preferences and assigned liberal profiles to the other side of the political spectrum. Also, AI cannot assess economic voting, which is the "punishment" of governments by the citizens. Artificial intelligence cannot take into account and analyze the benefits of the goals achieved by the current government for a specific individual. Also, the study confirmed the findings of previous studies that confirms Artificial intelligence favors liberal parties, which is reflected in the results of this study, where we see the most frequently proposed party, TS-LKD and Laisvės partija.

Grantor dc:publisher

Institutional Repository of Vilnius University

Year dc:date

2024

Author dc:creator

• Masiokaitė, Beata,

Statement dc:rights

• info:eu-repo/semantics/openAccess

Language dc:language

lit

Repository record dc:identifier

https://repository.vu.lt/VU:ELABAETD210582038&prefLang=en_US

OAI identifier oai:identifier

oai:vu.lt:elaba:210582038