引導式測驗 · 中等 等級
如何使用人工智慧測驗建構世界
透過測試虛構規則、追蹤連續性、將研究與發明分開以及保留作者的選擇來實踐人工智慧輔助的世界建構。
問題 1 的 8
A writer establishes that a heat charm becomes dangerous after repeated use. What is a productive AI request?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- A writer establishes that a heat charm becomes dangerous after repeated use. What is a productive AI request?
- Brainstorm possible effects on daily life, then check which serve the story
- Decide that the rule no longer applies
- Invent a real-world cultural tradition that matches the magic
- Add every suggested consequence to the novel
- How should a writer keep worldbuilding brainstorms from becoming accidental canon?
- Mix all ideas into one unmarked document
- Separate established facts from possibilities in a source-of-truth record
- Assume the latest model output overrides the draft
- Delete the manuscript after outlining
- A continuity tool flags a travel-time conflict with an earlier chapter. What should the writer do?
- Change the entire map immediately
- Assume the model found a definite error
- Inspect the cited passages and verify whether the conflict is real
- Remove all travel scenes
- How much worldbuilding should a story include?
- Enough to document every historical event
- Exactly one encyclopedia per book
- More detail than the plot can use
- Enough to serve the story and answer relevant reader questions
- A fantasy society is inspired by a real culture. What should the writer do?
- Research relevant sources and context rather than treating AI output as evidence
- Rely on AI-generated customs as factual
- Combine unrelated practices because they sound exotic
- Assume one person represents everyone in that culture
- A new scene intentionally changes an established world rule. What is a responsible next step?
- Hide the change from readers and collaborators
- Record the revision and check which earlier scenes need updating
- Assume the continuity checker will repair the whole series
- Delete the source-of-truth notes
- What should a writer include in a compact world bible?
- Every generated possibility as a confirmed fact
- Only character names, never dates or rules
- Established details, limits, relationships and unresolved questions
- Unverified summaries of real cultures
- What does using AI to brainstorm fictional consequences establish?
- That every proposed effect must occur
- That the world is historically accurate
- That readers will accept every rule
- Nothing by itself; the author evaluates each idea against canon and story purpose