Chapter 18 Quiz: Generative AI — Multimodal

Chapter 18 Quiz: Generative AI — Multimodal

Multiple Choice

Question 1. Which of the following best describes how a diffusion model generates an image?

(a) It retrieves the most similar image from its training database and modifies it to match the prompt.
(b) It learns to reverse a noise-addition process, iteratively refining random noise into a coherent image guided by a text description.
(c) It assembles an image by combining fragments of existing images from its training data.
(d) It converts the text prompt into a set of pixel coordinates and renders the image directly.

Question 2. In the opening scenario, Professor Okonkwo shows a third AI-generated image that contains a product feature that does not exist on the actual item. This example illustrates which key challenge of generative AI?

(a) The aesthetic quality gap — AI images do not look realistic enough.
(b) The accuracy gap — AI can generate beautiful but factually incorrect images.
(c) The cost gap — AI-generated images are still too expensive for commercial use.
(d) The format gap — AI cannot generate images in the required file formats.

Question 3. Which of the following is NOT a current capability of AI-powered image editing tools?

(a) Inpainting — replacing selected regions of an image with AI-generated content
(b) Outpainting — extending an image beyond its original boundaries
(c) Guaranteed accuracy — ensuring that edited images precisely match product specifications
(d) Background removal and replacement at near-perfect quality

Question 4. According to the chapter, a 2024 study found that listeners could distinguish AI-generated speech from human speech approximately what percentage of the time?

(a) 32 percent
(b) 52 percent
(c) 72 percent
(d) 92 percent

Question 5. Tom describes AI code generation tools as "writing code the way an intern does." Which of the following best captures the business implication of this observation?

(a) Code generation tools should not be deployed because they produce low-quality code.
(b) Code generation tools increase output speed but require senior developer review, shifting the required skill mix rather than eliminating roles.
(c) Code generation tools are only useful for junior developers and have no value for experienced engineers.
(d) Code generation tools will completely replace junior developers within two years.

Question 6. In the Getty Images v. Stability AI lawsuit, what is the core legal question?

(a) Whether AI-generated images are of sufficient quality to compete with professional photography
(b) Whether training an AI model on copyrighted images without permission constitutes copyright infringement
(c) Whether Getty Images owns the copyright to all images on the internet
(d) Whether Stability AI's pricing model undercuts Getty's business unfairly

Question 7. According to US Copyright Office guidance (2023-2025), which of the following statements about AI-generated content is correct?

(a) All AI-generated content is automatically copyrightable by the person who wrote the prompt.
(b) Purely AI-generated content without significant human creative input is not copyrightable.
(c) AI-generated content is copyrightable only if the AI model was trained exclusively on public domain data.
(d) AI-generated content is never copyrightable under any circumstances.

Question 8. What is the C2PA standard?

(a) A government regulation requiring all AI-generated content to be labeled as such
(b) A content provenance standard that embeds cryptographic metadata to establish a verifiable chain of how content was created and modified
(c) A quality assessment tool that rates AI-generated images on a scale from 1 to 10
(d) A copyright registration system for AI-generated works

Question 9. In Athena's AI image generation pilot, what was the approximate error rate for factual inaccuracies (wrong product features) in the AI-generated images?

(a) 2 percent
(b) 5 percent
(c) 11 percent
(d) 25 percent

Question 10. Which of the following best describes the "AI-assisted, human-verified" framework proposed in the chapter?

(a) AI generates content autonomously; humans check only for legal compliance before publishing.
(b) Humans create all content; AI is used only for distribution and analytics.
(c) AI generates content under human direction; humans verify accuracy, brand consistency, and legal compliance before publishing.
(d) AI and humans work independently, and the better output is selected for publication.

Question 11. Which deployment model for generative AI is BEST suited for a healthcare organization that must comply with strict data privacy regulations and cannot send patient data to third-party APIs?

(a) API-based services from a major cloud provider
(b) Fine-tuned model hosted by the AI vendor
(c) Self-hosted open-source model on the organization's own infrastructure
(d) Any deployment model, since all AI providers are HIPAA-compliant

Question 12. NK observes: "If every brand uses the same AI tools, don't we all start looking the same?" Which strategic concept does this observation most directly relate to?

(a) First-mover advantage in AI adoption
(b) The commoditization of creative execution and the increasing value of creative direction
(c) The declining cost of AI inference
(d) The network effects of platform-based AI services

Question 13. Which of the following represents a correct sequence in the "new creative workflow" described in the chapter?

(a) AI generates concepts → Human selects → AI distributes → Human reviews
(b) Human defines strategy → AI generates concepts → Human selects and refines → AI produces at scale → Human reviews for quality
(c) AI analyzes market data → AI generates concepts → AI produces at scale → Human reviews → AI distributes
(d) Human generates concepts → AI selects the best → Human produces → AI reviews quality

Question 14. A 2023 Stanford study on AI-assisted coding found that developers using AI coding assistants:

(a) Produced code with fewer security vulnerabilities than those coding without assistance
(b) Produced code with more security vulnerabilities, likely due to false confidence in AI-generated code
(c) Produced code with the same number of security vulnerabilities regardless of AI assistance
(d) Refused to use AI coding assistants due to security concerns

Question 15. Which of the following is the MOST accurate characterization of the current state of AI video generation (as of 2026)?

(a) AI-generated video is indistinguishable from professionally produced video for all commercial purposes.
(b) AI video generation produces impressive short clips but faces significant limitations in duration, consistency, directorial control, and resolution for most commercial applications.
(c) AI video generation is not commercially viable for any purpose and will not be for at least a decade.
(d) AI video generation has fully replaced stock video services and is beginning to replace professional videography.

Short Answer

Question 16. Explain why Lena Park advises businesses to "build your content strategy on the assumption that the rules will get stricter, not looser" regarding AI-generated content and intellectual property. Provide two specific reasons supporting this advice.

Question 17. Describe three specific business risks that deepfakes pose to corporations (beyond general societal harms). For one of these risks, outline a preventive measure.

Question 18. The chapter states that "verification is as important as generation." Using the example of Athena's product photography pilot, explain why this is true and describe the three-step verification pipeline NK implemented.

Scenario Analysis

Question 19. A fashion retailer is considering replacing its entire product photography operation ($500,000 annual budget, team of three photographers) with AI-generated imagery. Based on the frameworks in this chapter, identify:

(a) Two potential benefits of this approach
(b) Two potential risks
(c) What the chapter would recommend as a more balanced approach, and why

Question 20. Your company's CEO forwards you an article claiming that "AI will eliminate 80% of creative jobs within five years." Using evidence and frameworks from this chapter, draft a three-paragraph response that provides a more nuanced assessment. Address both the displacement and augmentation narratives.

Answer key is available in Appendix B.