Verification Capacity
Verification capacity is the skill of being a smart editor rather than a passive reader. Just as you wouldn't trust a stranger's directions without checking a map, this is the habit of double-checking AI work, spotting its mistakes, and adding your own common sense to make sure the final result is actually correct.
This is the active competency required to manage AI-generated output by shifting from a mindset of consumption to one of critical oversight. It involves maintaining enough subject-matter expertise to identify hallucinations, knowing when the AI’s logic is insufficient, and resisting the urge to defer to the machine's authority, thereby ensuring the output meets professional standards.
The competency that performs epistemic labour, resting on three things: epistemic depth (an 'oracle' to check against), metacognitive calibration (when to trust the machine versus yourself), and the psychological disposition to engage rather than defer. Grounded in epistemic vigilance (Sperber et al., 2010). Why AI raises the cognitive bar except on borrowed-output floor tasks. (C4AIL.)