Викторина, связанная с руководством · Средний Уровень

Как сделать проекты искусственного интеллекта человечными

Пересматривайте проекты ИИ, чтобы обеспечить ясность, доказательства и личный голос, не полагаясь на уклонение от обнаружения или искажение авторства.

Вопрос 1 из 8

Before rewriting a generic AI draft, what should the editor identify?

Предварительный просмотр каждого вопроса8 вопросы

Ответы и пояснения появляются по мере прохождения теста.

  1. Before rewriting a generic AI draft, what should the editor identify?
    • Identify what is missing for the audience, such as evidence or a concrete example
    • Add random spelling mistakes
    • Use a detector to choose the next sentence
    • Insert an invented personal story
  2. Why replace a vague claim with a specific verified example?
    • It makes every claim legally safe
    • It gives the reader concrete information that can be checked
    • It guarantees the author sounds casual
    • It reduces the need for sources
  3. What should a writer do with an invented statistic in a draft?
    • Сохраните, если это звучит правдоподобно
    • Change the number until it seems realistic
    • Remove it or replace it with verifiable evidence
    • Attribute it to an unnamed source
  4. What outcome should guide editing an AI draft?
    • Reduce the detector score
    • Add imperfections to imitate a person
    • Change every sentence length at random
    • Improve meaning, clarity, evidence and fit for the audience
  5. How should a writer use detector-evasion tools?
    • Avoid using them to disguise authorship or bypass a policy
    • Use them to prove the facts are accurate
    • Rely on them to establish originality
    • Use them instead of checking citations
  6. What does reading a draft aloud help reveal?
    • Whether its sources are licensed
    • Stiff phrasing, repetition or awkward sentence rhythm
    • Whether every fact is true
    • Whether Google will rank it
  7. What should a writer do with AI suggestions that add unsupported facts?
    • Keep them if they improve flow
    • Hide them in a footnote
    • Reject or verify them before inclusion
    • Assume the system checked the source
  8. What should guide AI use in school or workplace writing?
    • The model’s default disclosure preference
    • A detector’s classification alone
    • Whether the draft sounds casual
    • The relevant institution or publisher’s attribution policy