Metaphors and algorithms – embodied chatbots and their perception by users

The subject matter of the paper is based on the paradigm of conceptual metaphor, which assumes that metaphor connects reason and imagination and is not only a key component of everyday language use, but above all concerns thinking and cognition and influences the way we perceive and act. In this approach, conceptual projection is performed, i.e., the conceptual structure is transferred from the source domain to the conceptual structure of the target domain. This allows us to refer to experiences (physical and mental) related to the source domain in order to better understand the conceptual concept of the target domain. In this sense, metaphor has been the basis for designing information spaces and building user experiences for many years. The aim of the paper is to find an answer to the question of whether imagery schemes built on the basis of metaphors are reflected in communication with AI assistants (chatbots), as well as to examine whether the metaphor used affects the user’s perception of the chatbot’s warmth and competence, and overall satisfaction. The thesis of the presentation is based on the “Great Chain of Being” model, according to which AI assistants were given the characteristics of the following metaphorical figures: god, human, animal, plant, physical object, and assistant without characteristics. The research was conducted using eye-tracking, surveys, and in-depth interviews, and ChatGPT-4 and Dalle-3 technologies were used.