// library · pillar 05
Judgment & limits
Knowing when not to; failure modes, over-reliance.
16 modules
· all six pillars →
2026·09·17 Leave out who said it 2026·09·10 It copied without the quote marks 2026·09·07 What's missing doesn't look wrong 2026·08·25 Judge first, then ask: the order that beats double-checking 2026·08·24 The harsh verdict gets a pass 2026·08·18 Ask for a hint, not the answer: taking the help without losing the skill 2026·07·30 A share link is a public URL: what the 'share chat' button actually does 2026·07·28 The verification divide: why the same AI helps the best and hurts the rest 2026·07·23 Inherited capture: on someone else's call, the note-taker runs on their defaults 2026·07·15 Verify the request, not the voice: why spotting the deepfake is the wrong defence 2026·07·09 The give-up reflex: what ten minutes of AI help does to your persistence 2026·07·02 It agrees because you pushed back: spotting sycophancy before it changes your mind 2026·06·24 The summary that ate the caveat: what AI quietly drops when it condenses your documents 2026·06·22 The account is not the artifact: judging AI work you didn't watch get made 2026·06·15 Citation roulette: the verification habit that keeps you out of the AI-hallucination court database 2026·06·08 Why it makes things up: hallucination is a feature of the scoring, not a glitch you can prompt away