Subjective Interchangeability of Artificial Intelligence

This is the English version of the original Italian text published on Zenodo (DOI: https://doi.org/10.5281/zenodo.18291878). This text introduces the concept of the subjective interchangeability of artificial intelligence. The thesis is that AI represents the first technical artifact in the history of humanity that is interchangeable with a subject and, as such, can be held functionally responsible. Not because it is actually cognitive or embodied, but because it produces outputs that, in form and function, are phenomenologically similar to the outcomes of human cognition. This similarity makes plausible the attribution to AI of judgments, evaluations, and decisions, allowing it to occupy the place of the subject without the substitution appearing false. The text clarifies that such interchangeability should not be understood in a linguistic or legal sense, but as an operational possibility of cognitive attribution. It also demonstrates the discontinuity with previous technologies, which, while automating procedures, remained manifestly non-cognitive and not subjectivizable. The proposed thesis is ontological in a minimal sense: it identifies a new condition of possibility that precedes the analysis of the social, political, or decisional effects of artificial intelligence.