SELF-PERCEPTION AND SUBJECTIVITY: CHATGPT'S PERSPECTIVE ON ITSELF

Abstract

 The creation of language models, such as ChatGPT, represents one of the most revolutionary technological innovations in the field of artificial intelligence. These models not only expand the capacities for interaction between humans and machines but also raise fundamental questions about the nature of cognition, consciousness, and identity in artificial systems. Exploring the subjectivity and perspective that a model like ChatGPT can provide regarding its own creation offers a unique opportunity to investigate the boundaries between technology and the essence of human experience. Based on this, the central objective of this research is to examine how ChatGPT refers to itself and the processes involved in its development, aiming to understand how the model articulates an "identity" and a "perspective" of its own. The methodology adopted is a narrative approach, in which interactions with ChatGPT will be systematically recorded and subjected to qualitative analysis. Data was collected through specific questions directed at the model, focusing on its "life history" and aspects of its creation. Life history is an essential procedure in narrative research, allowing the collection of narratives that aid in understanding complex phenomena. This approach allows the identification of discursive patterns that simulate subjectivity and self-perception in the model, offering a detailed view of its responses. The results reveal how ChatGPT, even as a non-conscious AI model, can construct narratives that mimic human subjectivity, while simultaneously highlighting the model’s intrinsic limitations in representing a truly personal perspective. These findings may contribute to broader discussions on the ethical and philosophical implications of personifying artificial intelligence technologies, as well as on the meaning of subjectivity in non-human systems.