Week 1 Quiz — Welcome to the AI Revolution

Course: Using Artificial Intelligence (AI 101) · Silver Oak University (fictional sample) · Prof. Quinn
Covers: what generative AI is · the AI / generative AI / LLM / AGI vocabulary · the working mindset (general→specific, iterate) · generation vs. search · fluency ≠ truth
Format: 10 auto-graded items (multiple-choice, multiple-answer, matching, true/false) · 10 points (1 each) · allowed attempts: 1 · No AI on this quiz.

This is the human-readable quiz with its vetted answer key and one-line feedback. The import-ready Classic QTI 1.2 is in F-quiz-week-01-qti.xml (generated by a validated Python script — parses with 10 items, every single-answer item exactly one correct). Reminder: AI is not permitted on quizzes — this checks that you understand the Week 1 ideas.

Questions, key, and feedback

Q1 (MC). Generative AI is best described as —
- A. software that searches the internet and returns existing answers
- B. software that creates new content (such as text or images) in response to a request ✅
- C. a robot with human-level general intelligence
- D. a fixed database of pre-written answers
Feedback: Generative AI creates new content. It isn't a search engine (A), it isn't AGI/a robot (C), and it isn't a lookup table (D). The "generates" verb is the whole idea.

Q2 (MC). When you open a chatbot like ChatGPT, the large language model (LLM) is —
- A. the same thing as the chatbot app you opened
- B. the text-prediction engine inside the app that generates the replies ✅
- C. a search engine that the app queries for facts
- D. a team of humans who type the responses
Feedback: The chatbot is the app; the LLM ("the model") is the engine inside it. Keep the app and the engine separate — it's the most common Week 1 mix-up.

Q3 (Matching). Match each term to its correct description.
| Term | Correct description |
|---|---|
| Artificial intelligence (AI) | The broad field of computers doing tasks that once needed human intelligence |
| Generative AI | The slice of AI that creates new content (text, images, audio, code) |
| Large language model (LLM) | The text-prediction engine inside a chatbot |
| AGI (artificial general intelligence) | A hypothetical future AI that could do any human intellectual task |
Feedback: Nested set: AI (umbrella) → generative AI (makes content) → LLM (the engine) … and AGI is the not-yet-real "can do anything" goal. "AI is the field, generative AI makes things, the LLM is the engine, AGI is the sci-fi goal."

Q4 (MC). Which statement about AGI (artificial general intelligence) is correct?
- A. AGI is what powers today's chatbots
- B. AGI is a hypothetical future AI that does not exist today ✅
- C. AGI is just another name for a large language model
- D. AGI means an AI that can generate images instead of text
Feedback: Today's tools are powerful but narrow; AGI does not exist yet. Don't confuse the impressive-but-narrow tools you use with the hypothetical "do-anything" AGI.

Q5 (Multiple answer — select all that apply). Which of the following are TRUE about how today's large language models work?
- A. They generate text by repeatedly predicting the next chunk of text ✅
- B. They can be fluent and factually wrong at the same time ✅
- C. They look up every answer in a verified database of facts
- D. They possess human understanding and beliefs
- E. Their output should be verified, especially specific facts, numbers, and citations ✅
Feedback: True: A (next-token prediction), B (confidently wrong), E (verify). C is false — it generates, it doesn't look up a fact database. D is false — no understanding or beliefs behind the text.

Q6 (MC). You ask an AI for help and its first answer is too generic. Based on the Week 1 mindset, the best next move is to —
- A. give up, because the AI clearly cannot help
- B. retype the exact same prompt and hope for a better result
- C. steer it with a more specific follow-up (add context, ask for a format or examples) ✅
- D. assume the generic answer is the best the AI can do
Feedback: The first answer is a draft. Iterate: general → specific, with follow-ups. That back-and-forth is where the value is.

Q7 (MC). Which revision most improves the weak prompt "help me with my essay"?
- A. "please help me with my essay, thank you"
- B. "HELP ME WITH MY ESSAY" (typed in capital letters)
- C. "I'm writing a 3-page argumentative essay for a first-year college class on whether remote work helps productivity. Give me three possible thesis statements." ✅
- D. "help me with my essay right now"
Feedback: The strong fix adds context (who/what), a clear goal, and a format (three thesis options). Politeness (A), capitals (B), and urgency (D) don't give the AI what it needs.

Q8 (True / False). If an AI's answer is fluent, confident, and specific, that means it is factually correct.
- True
- False ✅
Feedback: False. Fluency ≠ truth. An LLM generates plausible-sounding text, so it can be confident, specific, and wrong all at once — which is exactly why you verify.

Q9 (MC). A student asks a chatbot to "write a haiku about my cat." The chatbot is mainly —
- A. finding an existing haiku that was already written about a cat
- B. generating brand-new lines of text based on patterns it learned ✅
- C. copying a haiku word-for-word from a database
- D. proving it understands poetry the way a human poet does
Feedback: It generates new lines (B) — that's the difference from search (A/C). And producing a poem isn't proof of human-style understanding (D).

Q10 (MC). The idea "the machine has no brain — use your own" means that —
- A. AI is useless, so you should do everything yourself
- B. because the model predicts text rather than knowing things, the human must set the goal, judge the output, and verify the facts ✅
- C. you should never use AI for anything that matters
- D. the AI is always right, so you do not need to think
Feedback: The point isn't that AI is useless (A/C) or infallible (D) — it's that you stay the judge: set the goal, evaluate, and verify. The tool drafts; you decide.

Answer key (quick reference)

Blueprint & item-bank note

All 10 items are tagged course=AI101 · week=1 · objective=1 and deposited into the item bank for future per-term ($39) regenerations. Distractors target the week's classic misconceptions (AI = search engine; the app is the model; AGI is here; fluent = true; magic words).

Quality gate (self-checked)

• Structure: 10 items, 1 point each; types = 7 multiple-choice + 1 matching + 1 multiple-answer + 1 true/false.
• Single-answer integrity: every MC and the true/false item has exactly one correct option; the matching item pairs one-to-one; the multiple-answer item keys A, B, E (and requires C and D to be left unselected).
• Product-accuracy gate: PASS. Real products named factually (ChatGPT and the LLM-vs-app distinction); no fabricated features; conceptual claims (generation vs. search, next-token prediction, fluency ≠ truth) are accurate and non-controversial.
• QTI parse confirmation: F-quiz-week-01-qti.xml parses as imsqti_xmlv1p2 with 10 items; each single-answer respcondition sets SCORE = 100 on exactly one option; the matching item's partial-credit blocks sum to 100; the multiple-answer item requires the exact A/B/E set.

Canvas placement block

canvas_object    = Quizzes::Quiz
title            = "Week 1 Quiz — Welcome to the AI Revolution"
assignment_group = "Quizzes"
points_possible  = 10
grading_type     = points
available_from_offset_days = 0
due_offset_days  = 6
published        = true
allowed_attempts = 1
shuffle_answers  = true
ai_permitted     = false
provenance       = "~ Prof. Quinn's edition · Fall 2026 · built with thecoursemaker.com"