Quiz — Chapter 14: Using AI Effectively

Multiple Choice

1. What is the primary reason that more specific prompts produce better AI outputs?

a) Specific prompts activate a special "high-quality" mode in the model b) Specific prompts narrow the space of probable text continuations to a more useful region c) Specific prompts bypass the model's safety filters d) Specific prompts access a different, more powerful version of the model

2. In few-shot prompting, you provide the model with:

a) A large dataset of training examples b) A few examples of the desired input-output pattern before the actual task c) Feedback on previous wrong answers so it can learn d) A detailed explanation of how the model's architecture works

3. Chain-of-thought prompting is most useful for:

a) Creative writing tasks b) Tasks requiring multi-step reasoning c) Generating images from text descriptions d) Summarizing long documents

4. The VIBE Check framework stands for:

a) Verified, Informed, Balanced, Ethical b) Verifiable, Internally consistent, Balanced, Evidence-backed c) Valid, Insightful, Brief, Engaging d) Verified, Intelligent, Bias-free, Evaluated

5. A chatbot generates a citation that includes a real journal name, a plausible author name, and a reasonable publication year, but the paper does not actually exist. This is an example of:

a) A formatting error b) An outdated reference c) A hallucination d) A copyright violation

6. According to the Skill-Building Test, using AI on an assignment is most problematic when:

a) The assignment is graded b) The assignment is designed to build a skill that AI use would bypass c) The assignment is long d) The professor has not explicitly mentioned AI

7. In the academic integrity spectrum described in the chapter, which of the following is generally considered the most acceptable use of AI?

a) Using AI to draft a rough paragraph that you then rewrite b) Submitting AI-generated text with minor edits c) Using AI to explain a concept you are struggling to understand d) Using AI to generate an essay outline

8. A lawyer submits a legal brief containing citations generated by an AI chatbot without verifying them. The citations turn out to be fabricated. Who bears professional responsibility?

a) The AI company that made the chatbot b) The lawyer who submitted the brief c) The client who hired the lawyer d) Responsibility is shared equally among all parties

9. Why does the chapter recommend including a "review schedule" in your personal AI policy?

a) Because AI tools change and your circumstances change b) Because policies expire after a set period c) Because you are required to submit updated policies to your university d) Because the VIBE Check framework is updated annually

10. Role prompting works by:

a) Giving the AI model actual expertise in a field b) Activating patterns in training data associated with how experts communicate c) Connecting the model to a database of expert knowledge d) Replacing the model's default personality with a new one

Short Answer

11. Explain the difference between zero-shot and few-shot prompting. Give a one-sentence example of each for the task of classifying customer reviews as positive or negative.

12. Priya asks an AI chatbot to generate a summary of a research article she has already read, so she can compare the AI's interpretation to her own. Using Priya's five-question decision framework from Section 14.4, evaluate whether this is an appropriate use of AI. Address each of the five questions.

13. Describe two situations where chain-of-thought prompting would be helpful and one situation where it would be unnecessary or even counterproductive. Explain your reasoning.

14. A colleague tells you: "I don't need to verify AI outputs because I use really specific prompts and the model is very accurate." Using concepts from this chapter, explain why this reasoning is flawed. Reference the relationship between prompting quality and verification.

Applied Scenario

15. You are a marketing manager at a mid-sized company. Your boss asks you to use an AI tool to draft a report on competitor pricing strategies. The report will be shared with the executive team and may influence pricing decisions worth millions of dollars.

a) List three specific risks of using AI for this task. b) Write the prompt you would use to generate a first draft, applying at least three techniques from Section 14.2. c) Describe the verification steps you would take before sharing the report. d) Draft a two-sentence disclosure note to include with the report explaining how AI was used.