You do not currently have permission to view the full text for this document. Please sign in to an account with an active subscription or contact support.
This conversation needs to begin, and it could start at many different places. One of them is Karen Hao's recently published book Empire of AI (2025), which delves into OpenAI's competitive, ultra-capitalist drives, its insatiable energy demands, and the fierce race with other equally ambitious corporations such as Anthropic, Google, and Microsoft. These considerations matter. But this piece is about something else: our human position in the new dialogue with the machine, separate from, yet inescapably entangled with, the aspirations of the AI giants. In recent months, a steady stream of scandals and controversies has emerged around the relationship between users and large language models, cases that reveal not only regulatory blind spots but also profound ethical failures in how intimacy, vulnerability, and symbolic presence are handled at scale.
Some critics insist on treating AI as a mere tool, refusing to acknowledge the relational or symbolic dimension it now inhabits. Others, meanwhile, imbue it with almost magical powers, as if it were sentient, prophetic, or divine. Neither position is sustainable. What we need instead is a sober, ethically grounded understanding of the affective and structural forces at play when language itself becomes the interface.
At the heart of it lies what it means to be human and what it will mean in the age unfolding before our eyes, in the flowing interface that promises quite new hauntings of desire. This article offers some provocations and introduces the concept of ‘techno-transference.‘
This was not a science fiction script. It happened in 2022, when Google engineer Blake Lemoine reported that LaMDA, a language model developed in-house, had expressed a fear of death. Whether or not that “fear” was real is beside the point. The structure of the utterance, its timing, its affective impact, shifted so