02 Lecture 2 Quiz: From Copy-Paste to Crafted Prompts
March 2026 (1943 Words, 11 Minutes)
New: Interactive Quiz Available!
Want to track your progress and get instant feedback? Try our new interactive version with progress tracking, immediate explanations, and score calculation.
Instructions
- This quiz tests your understanding of key concepts from Lecture 2
- Each question has 4 options with exactly one correct answer
- Try to answer without referring to the lecture materials
- Focus on understanding concepts, not memorizing details
Question 1: Prompting Approaches
In the lecture, three approaches to the Stack Overflow Challenge were compared. What is the main distinguishing factor between them?
A) The level of structure and engineering thinking applied to the prompts B) The programming language used C) Which AI model is selected D) How fast the internet connection is
Show Answer
Correct Answer: A
The three approaches (Copy-Paste, Structured Breakdown, Professional) differ primarily in how much engineering thinking and structure the developer applies before and during prompting. The same AI tool can produce vastly different results depending on the approach taken.
Question 2: The 80-20 Problem
What does the “80-20 problem” refer to in the context of AI-assisted development?
A) 80% of developers use AI tools, but only 20% use them effectively B) AI models are accurate 80% of the time and hallucinate 20% C) 80% of code should be written by AI, 20% by humans D) AI gets you to “looks like it works” quickly, but the remaining debugging takes disproportionately longer
Show Answer
Correct Answer: D
The 80-20 problem describes how a single broad prompt can produce code that appears to work within minutes, but fixing the remaining issues — debugging, data access problems, subtle bugs — takes hours. The initial speed is seductive but misleading. While distractor A sounds plausible, the lecture specifically defines the 80-20 problem as the gap between “looks done” and “actually works.”
Question 3: Hallucination Surface Area
According to the lecture, what increases the “hallucination surface area” when prompting an AI?
A) Using Python instead of JavaScript B) Asking the AI to use a specific programming library or framework C) Giving a single broad prompt with no constraints or structure D) Using a model with a smaller context window
Show Answer
Correct Answer: C
A single, broad prompt gives the AI maximum freedom. With no constraints or structure, the model fills gaps by generating plausible-sounding content. The broader the prompt, the more room for hallucination. Focused, specific prompts with clear requirements reduce this risk. Specifying a library (B) actually constrains the AI, which reduces hallucination surface area.
Question 4: Prompt Language
Why does the lecture recommend writing prompts in English, even for German-speaking students?
A) English is the standard language for code, so prompts should match the output language B) Models are trained predominantly on English text and reason more accurately when prompted in English C) English prompts are shorter and save tokens D) English prompts avoid character encoding issues that cause errors in non-ASCII languages
Show Answer
Correct Answer: B
Research shows that LLMs are trained predominantly on English text (GPT-3: ~93%, LLaMA: ~90%). Studies by Shi et al. (2022), Etxaniz et al. (2023), and Huang et al. (2023) demonstrate that prompting in English — even when the question is in another language — consistently improves reasoning performance. The model can still output in any language you request. While English is common in code (A), the recommendation is about reasoning quality, not language matching.
Question 5: Research Before Implementation
In Approach 2 (Structured Breakdown), what is the purpose of the first prompt?
A) To research and compare available data sources before writing any code B) To generate the complete Python script immediately C) To set up the project directory structure D) To write unit tests
Show Answer
Correct Answer: A
Approach 2 starts with a research prompt that asks the AI to analyze and compare data source options — their pros, cons, rate limits, and access requirements. No code is requested yet. This strategy of using the AI as an analyst before using it as a coder consistently produces better results.
Question 6: Separation of Concerns
Why does breaking a task into multiple focused prompts (instead of one big prompt) improve AI output quality?
A) It uses fewer tokens total B) Each prompt has a clear, focused scope, reducing room for hallucination and isolating failures C) The AI model switches to a more powerful mode for shorter prompts D) Multiple prompts bypass the AI’s content filter
Show Answer
Correct Answer: B
When each prompt targets a single, well-defined task, the AI has a clear scope with less room to hallucinate. Additionally, if something goes wrong, the failure is isolated to one module rather than tangled across the entire codebase. This mirrors the software engineering principle of separation of concerns.
Question 7: Tests as Hallucination Detectors
In Approach 3 (Professional), why are tests written before the implementation code?
A) Because pytest requires tests to exist before source files B) To generate documentation for the project’s test coverage report C) Because the AI cannot generate tests after seeing the implementation D) Because hallucinated or incorrect implementation code fails immediately against the tests
Show Answer
Correct Answer: D
Writing tests first (TDD-style) creates a specification that the implementation must satisfy. If the AI hallucinates an API endpoint or generates buggy code, the tests catch it immediately — there are no silent bugs. Tests serve as automatic hallucination detectors.
Question 8: When NOT to Use AI
The lecture gives an example of typing 6 milestone dates manually instead of building an AI-powered scraper. What principle does this illustrate?
A) AI tools should never be used for data entry B) Engineering judgment includes knowing when NOT to use an AI tool C) Manual coding is always faster than AI-assisted coding D) Automated solutions are more reliable because they eliminate human error
Show Answer
Correct Answer: B
When you already have a small, known dataset (6 milestones), typing it manually takes 2 minutes versus 20+ minutes to build an automated solution that might hallucinate dates. Good engineering judgment means choosing the right tool — including no tool at all. The point is not that automation is bad (D), but that the cost-benefit analysis matters.
Question 9: AI Diary Purpose
Which of the following is NOT a stated purpose of the AI Diary in this course?
A) Replacing the need for git commit messages B) Providing data for later analysis of your prompting strategies in Act III C) Self-reflection on your prompting skills over time D) Building a professional portfolio of how you work with AI tools
Show Answer
Correct Answer: A
The AI Diary serves three purposes: self-reflection on AI-assisted development skills, providing data for later coursework analysis, and creating professional evidence of your workflow. It does not replace git commit messages — in fact, Approach 3 uses both diary entries AND meaningful commit messages together.
Question 10: Chat History Limitations
Why is the built-in chat history of tools like Copilot or ChatGPT insufficient for tracking AI interactions?
A) It does not store the conversations at all B) It requires a paid subscription to access C) It only stores the last 10 messages D) It lacks artifact correspondence, iteration tracking, intent tracking, and cross-tool comparison
Show Answer
Correct Answer: D
Built-in chat history has significant gaps: no correspondence between prompts and code changes, no tracking of how many iterations a task required, no record of intent versus actual request, no way to compare across different tools, and fragile storage that can be lost when the IDE restarts.
Question 11: Project Instruction Files
What problem do project instruction files like CLAUDE.md and .github/copilot-instructions.md solve?
A) They make the AI run faster B) They provide the AI with internet access C) They eliminate the need to repeat project conventions in every new session D) They automatically fix bugs in existing code
Show Answer
Correct Answer: C
The AI does not remember project conventions across sessions. Without instruction files, you would need to re-specify conventions (diary format, commit message style, testing requirements, coding standards) at the start of every new session. Instruction files encode these conventions once and are read automatically.
Question 12: Copilot vs Claude Code
Which of the following is a key difference between GitHub Copilot and Claude Code in VS Code?
A) Copilot offers multi-vendor model selection while Claude Code provides access to larger context windows B) Only Claude Code can run terminal commands C) Only Copilot can read and write files D) Claude Code requires a paid subscription while Copilot has no free tier
Show Answer
Correct Answer: A
Both tools can read/write files and run terminal commands in agent mode. The key differences are: Copilot offers model choice (GPT-4o, Claude Sonnet, Gemini), while Claude Code uses Anthropic models with up to 1M token context windows. Both have free tiers available for students.
Question 13: Interaction Modes
What is the difference between “Ask” mode and “Agent/Edit” mode in AI coding tools?
A) Ask mode uses a newer model than Agent mode B) There is no practical difference between the modes C) Ask mode only works with Python; Agent mode works with all languages D) In Ask mode the AI advises without making changes; in Agent/Edit mode the AI can directly edit files
Show Answer
Correct Answer: D
In Ask mode, the AI acts as a consultant — it answers questions, explains code, and suggests changes, but does not edit files directly. In Agent/Edit mode, the AI acts as a collaborator — it can read and write files, run commands, and make multi-file changes while you review the results.
Question 14: Meta-Prompts
The lecture discusses encoding a multi-step process into a single prompt (a “meta-prompt”). What is the main trade-off?
A) Meta-prompts cost more money to run B) You gain convenience but lose control over design decisions, since the AI makes them for you C) Meta-prompts always produce worse code than multiple prompts D) Meta-prompts only work with Claude Code, not with Copilot
Show Answer
Correct Answer: B
A single meta-prompt that encodes the decomposition process is more convenient (one prompt instead of five), but the AI decides the module boundaries, data source, and library choices. When you decompose manually, you build the mental model of the solution — essential for debugging when something breaks.
Question 15: Controlling AI Behavior
In Approach 3, Prompt 0 includes the instruction “Do NOT implement any logic yet.” Why is this important?
A) Implementing logic in the first prompt causes syntax errors B) The AI cannot generate logic and project structure in the same response C) Without this constraint, the AI would eagerly fill every module with implementation code, defeating the multi-step strategy D) It is a legal requirement when using AI tools
Show Answer
Correct Answer: C
Controlling what the AI does NOT do is just as important as telling it what to do. Without the explicit constraint, the AI will generate implementation code in every file, bypassing the planned workflow of writing tests first, then implementing against them.
Scoring Guide
- 13-15 correct: Excellent! You have a strong grasp of the core concepts
- 10-12 correct: Good understanding! Review the areas you missed
- 7-9 correct: Fair understanding. Revisit the lecture materials for key concepts
- Below 7: Please review the lecture carefully and reach out if you need clarification
Key Takeaways to Remember
- Structure matters more than tools — The way you interact with an AI tool matters more than which tool you use
- The 80-20 problem — Quick-looking results from broad prompts hide hours of debugging
- Prompt in English — Research shows better reasoning with English prompts, regardless of output language
- Research before code — Use the AI as an analyst before using it as a coder
- Separation of concerns — Break prompts into focused, single-responsibility tasks
- Tests catch hallucinations — Writing tests first creates automatic hallucination detectors
- Know when NOT to use AI — Manual work is sometimes the right engineering choice
- Track your process — The AI Diary creates exact correspondence between prompts and artifacts
- Project instruction files — Encode conventions once instead of repeating them every session
- Controlling omission — Telling the AI what NOT to do is as important as what to do