Ontological Invisibility

Ontological invisibility names the structural property of AI-collaborative authorship that distinguishes it from traditional editorial collaboration: the machine's contributions are invisible not because professional conventions hide them but because the collaboration produces an integrated output that cannot be decomposed into discrete human and machine components. When an author works with a human editor, the collaboration unfolds in stages: the author writes a draft, the editor responds with suggestions, the author revises, the editor comments again. Each stage is discrete and, in principle, traceable. The author can point to specific changes and say "my editor suggested this." When an author works with AI in real time—thinking through Claude, developing ideas in dialogue, receiving formulations before intentions have fully crystallized—the collaboration does not produce discrete stages. The assistance and the authorship arrive together, merged at the point of generation. The author who later tries to separate them faces not a problem of memory or record-keeping but a categorical impossibility: the thoughts were never cleanly separable, because they were produced in a cognitive space