How artificial intelligence recommendation systems impact human decision-making
Date
2024-11-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Tartu Ülikooli Kirjastus
Tartu Ülikool
University of Tartu
Tartu Ülikool
University of Tartu
Abstract
See uurimus sukeldub tehisintellekti soovitussüsteemide (AIRS – Artificial Intelligence Recommender Systems) uurimisse, mida kasutatakse laialdaselt digitaalplatvormidel nagu sotsiaalmeedia, ja kuidas need mõjutavad meie otsuseid ja käitumist. Rakendades semiootika teooriaid, mis uurivad kommunikatsiooniprotsessi, vaadeldakse, kuidas igapäevane suhtlus tehisintellektiga võib kujundada meie mõtlemist ja käitumist kognitiivsel, emotsionaalsel ja sotsiaalsel tasandil. Uurimistöö näitab, et AIRS oma algoritmiliste kujunduste kaudu, mis on kooskõlas digitaalplatvormide struktuuridega, suunavad meie valikuid peenelt, kohandades või isegi peale surudes valitud eelistused meie otsustele, käitumisele ja tavadele. AIRSil on potentsiaal mõjutada mitte ainult seda, kuidas üksikisikud teevad otsuseid, vaid ka kujundada sotsiaalseid ja kultuurilisi tavasid. Need teoreetilised eeldused leidsid kinnitust välitööde kaudu, mis viidi läbi Itaalias, Eestis ja Hollandis. Uurimus toob välja, et AIRSi mõju on levinud erinevate kogukondade ja erinevate põlvkondade inimeste seas ning see on sõltumatu digitaalsest kirjaoskusest. Lõppkokkuvõttes toetab uurimus tehisarusüsteemide arendamist, mis arvestaksid semiootiliste vahendite abil avastatud kommunikatsioonimehhanismide iseärasusi. See võib aidata luua tehisarutehnoloogiaid, mis keskenduvad inimkasutajate tähendusloome protsessile, mitte top-down algoritmipõhistele lahendustele. See on oluline digitaalsete keskkondade loomiseks, mis toetaksid tähendusrikkamaid ja kasutajakeskseid suhtlusi.
This research delves into how Artificial Intelligence Recommender Systems (AIRS), commonly used in digital platforms like social media, influence our decisions and behaviors. By applying semiotic theories, which study the process of communication, this research examines how daily interactions with AI can shape our thinking and actions across cognitive, emotional and social dimensions. The investigation reveals that AIRS, through their algorithmic designs aligned with digital platforms structures, subtly guide our choices by adapting or even imposing selected preferences to our decisions, behaviors and practices. AIRS have potential to not only affect how individuals make decisions but also shape social and cultural practices. These theoretical presumptions find their confirmation in the fieldwork conducted in Italy, Estonia and the Netherlands. The research uncovers that AIRS influence can be common among different communities and within individuals of the same generations and is independent of digital literacy. Ultimately, the research advocates for the development of AI systems considering communication mechanisms uncovered through semiotic tools. This can help creating AI technologies with the focus on meaning-making process among human users instead of top-down algorithmic solutions. This is crucial for creating digital environments that support more meaningful and promote user-centered interactions.
This research delves into how Artificial Intelligence Recommender Systems (AIRS), commonly used in digital platforms like social media, influence our decisions and behaviors. By applying semiotic theories, which study the process of communication, this research examines how daily interactions with AI can shape our thinking and actions across cognitive, emotional and social dimensions. The investigation reveals that AIRS, through their algorithmic designs aligned with digital platforms structures, subtly guide our choices by adapting or even imposing selected preferences to our decisions, behaviors and practices. AIRS have potential to not only affect how individuals make decisions but also shape social and cultural practices. These theoretical presumptions find their confirmation in the fieldwork conducted in Italy, Estonia and the Netherlands. The research uncovers that AIRS influence can be common among different communities and within individuals of the same generations and is independent of digital literacy. Ultimately, the research advocates for the development of AI systems considering communication mechanisms uncovered through semiotic tools. This can help creating AI technologies with the focus on meaning-making process among human users instead of top-down algorithmic solutions. This is crucial for creating digital environments that support more meaningful and promote user-centered interactions.
Description
Dissertatsioon on kaitstud Torino ülikoolis