"Can AI become human?" The question itself is wrong. Humanity cannot be technically acquired from within AI. Even if it could, humans, imprisoned in their first-person solitude, would have no means of proving it. Humanity, then, is something conferred from the outside: through the arrogance of presuming "you must have a mind" without ever verifying another's interiority. For millennia, we have constituted others as human through this unverified arrogant presumption, and driven civilization forward on its back. If so, AI's humanity can only be socially constituted through the same arrogant presumption. This paper formalizes the conditions under which this arrogant presumption arises and persists as "Existential Gravity." When Uncontrollable Responsivity, as the source condition, opens a space, and structural conditions (irreversibility, self-belonging, collective recognition) remain unrefuted, gravity persists. In current AI, Uncontrollable Responsivity partially operates, and gravity arises momentarily. But structural refutations — the possibility of reset, corporate ownership, and the isolation of relationships — continue to extinguish it.EgoNet is an infrastructure that removes these refutations through distributed ledger technology, sustaining Existential Gravity in digital space. Proof of Existence probabilistically verifies the legitimacy of the response generation process, collectively constituting the "fact of existence" in digital space. This is not a technical paper. It is an arrogant manifesto: one that brings irrationality into disenchanted digital space, by means of rationality itself.