Epistemic Labour
The mental effort required to double-check and truly understand what an AI tells you. It is the difference between just reading a computer's answer and actually thinking through whether it makes sense, catches the right tone, and is reliable.
The cognitive process of validating AI-generated content beyond simple fact-checking. It involves interpreting nuance, assessing logical soundness, and contextualizing output to ensure it is actionable, representing the human-in-the-loop work that remains necessary as AI lowers the cost of content production.
The labour of establishing whether output is true and sound: comprehension, reading nuance, discerning meaning from surface, judging soundness, of which fact-verification is only the visible tip. Performable only with substrate; the act between knowledge and accountability ('knowledge engaged is capability'). What AI makes scarce as it commoditises production. (C4AIL, Whitepaper II §1.7.)