The article analyzes visions of the future of artificial intelligence presented by key stakeholders within the context of sociotechnological narratives. The study demonstrates how diverse groups—including scientists, entrepreneurs, ethicists, policy experts, and media—construct narratives about AI development that shape both the practical implementation of the technology and regulatory decisions. Employing an Actor-Network Theory approach alongside a computational grounded theory and computer-assisted qualitative methodology, supported by natural language processing tools and generative coding, the research systematically analyzes numerous expert interviews sourced from the popular MLST YouTube channel, enabling the identification of distinct narrative patterns. The findings indicate that these narratives function as translation mechanisms, converting technological complexity into shared frameworks of understanding, which can concurrently contribute to the black-boxing process of AI. The conclusions emphasize that the future of artificial intelligence is not solely the outcome of technological progress but is the result of continuous negotiations among social and technical actors. Overall, the article makes a significant contribution to interdisciplinary discussions on the role of narratives in shaping AI development, highlighting the need for enhanced transparency and critical analysis of the processes that create stabilized yet hidden decision-making systems.