Why did Claude NOT learn from a book of explicit rules?
ARules are too short to fit in a training file
BRules cost more compute than examples
CRules fail novel situations; examples teach the pattern instead
DRules were tried but always returned wrong answers
Answer & Solution
Correct answer: C. Rules fail novel situations; examples teach the pattern instead
Card 4 of this block: a rulebook breaks down in novel situations; learning from millions of examples lets the model generalise the underlying pattern. Source: claude_constitution_ocr.md -- Anthropic favors cultivating good judgment over strict rules because rules often fail to anticipate every situation (p.4).
Related questions
A friend asks: what should you do first when an AI answer contains a specific date?Which type of claim carries the highest hallucination risk?What is the main reason AI sometimes produces confident but wrong answers?You told AI your name yesterday. In a new chat today, AI does not know it. Why?AI says an essay has 850 words. The real count is 1003. What does this show?You ask AI for today's cricket score. Why can't it answer reliably?What does AI do well when you ask it to compare two options?Which task is AI best suited for?