The article examines the mechanisms of speech representation of artificial intelligence (AI) as a media personality. The relevance of the study stems from the rapid integration of neural networks into media communication and ongoing discussions about the potential replacement of humans with AI. The research aims to identify the challenges and opportunities of such a replacement through analysis of the speech representation of anthropomorphic AI systems. Using the concept of the linguistic persona and author’s methodology, which extrapolates the category of personality from humans to AI, the study analyzes a corpus comprising over 1000 AI-generated posts and 1500 corresponding user comments. The analysis showed that the speech representation of AI is characterized by internal inconsistency. Anthropomorphic features (such as names, visual image, and communication styles) serve primarily pragmatic purposes aimed at increasing audience loyalty, rather than reconstructing an authentic linguistic persona. The key conclusion is the evidence that a full replacement of humans by AI in the media sphere is currently impossible due to neural networks’ lack of stable reflexive identity and genuine subjectivity.
Yulia Konyaeva (Mon,) studied this question.