Who am I talking to?: Memory of Chatbots and Personalized Chat Interaction

Abstract

Developed by OpenAI in 2022, ChatGPT is a chatbot based on a large language model that has evolved beyond generating human-like dialogue to become a versatile tool producing a wide range of content, from official documents to literary texts. Research indicates that users often feel as though they are communicating with a human during their interactions with artificial intelligence. In 2024, OpenAI introduced the “conversational memory” feature to ChatGPT, enabling the system to store and process past interactions with users. This innovation allows the chatbot to refer to previous conversations and generate more personalized and contextually appropriate responses. At the same time, it raises ongoing debates regarding its implications for human experience and agency. This study aims to examine the impact of ChatGPT’s conversational memory feature on user experience. Employing a qualitative approach, semi-structured in-depth interviews were conducted with 30 ChatGPT users, divided into two groups: those who use the conversational memory feature and those who do not. The findings reveal that participants who utilized the memory feature described their interactions as more personalized, consistent, and emotionally engaging. In contrast, participants who refrained from using the feature reported more transactional and utilitarian experiences, often citing concerns related to data privacy and user autonomy. The study highlights that chatbots are no longer perceived merely as technical tools; rather, they are seen as digital agents capable of addressing users’ emotional needs. This study also underlines the need for studies on human-artificial intelligence interaction, digital connectedness, and trust in technological systems.

Ähnliche Arbeiten