In 2023, technologies based on Artificial Intelligence (AI) became available to many users of smartphones and digital services all over the world. AI is accessible in the form of Voice Assistants (VAs), music recommendations, Virtual Assistants, and even psychotherapy-bots. Such spread of AI also influenced the interest of psychologists in the topic of human-AI Interaction (HAII). In our work, following the tradition of social representation research by Moscovici [Moscovici 2015], we consider AI as an object of social representation. Using the VA as a form of AI, we reconstruct the semantic field of VA in a group of its users and offer a description of the process of its formation. The study took place in three stages: free associations, interviews, and survey. In the interview, the sample consisted of 15 people, in free associations – 168 people (3,329 associations with the phrase “Voice assistant” were analyzed), in the survey using the semantic differential method – 800 people. As a result of the thematic and frequency analysis, the following results were obtained: • The semantic field at the present stage of the study of VA is positiveneutral; the associations are divided into two topics: about the intellectual abilities and emotional qualities of the VA. • With semantic differential analysis, three factors were separately obtained: usefulness, emotionality, and humor. It is assumed that it is in the category of those three factors that the perception of a particular VA is formed. • The process of forming social representations is outlined with the help of interviews. Two main stages in “communication” with the VA were identified: the image formation and interaction with the finished image. As future research it was proposed to continue studying the impact of the image of the VA within the current perception model on various features of interaction with it (e.g. within the framework of economic decisions)
Read full abstract