L1
·
Quiz
·
Lab
L2
·
Quiz
·
Lab
L3
·
Quiz
·
Lab
L4
·
Quiz
·
Lab
Module Test
Lesson 1 · AI & Writing

What Does AI Actually Do With Words?

From autocomplete on steroids to full paragraphs — how language models learned to write.
How does an AI turn a simple prompt into a full sentence — and should we call that "writing"?
📚 Note for Learners — This module is designed for middle and high school students. We'll explain every technical term as we go and connect big ideas to things you already know.

In 2016, Google researchers published a paper describing a system called Word2Vec. The system had been trained on billions of words from the internet, and when researchers asked it to complete the equation "king − man + woman = ?", the system answered: "queen." The AI hadn't been told what a queen was. It had simply absorbed the patterns of how humans use words — and the relationship between those words — from reading more text than any human ever could.

Prediction, Not Understanding

Here's the most important thing to understand about AI writing tools: they predict, they don't understand. A language model like GPT-4, Claude, or Gemini reads your prompt and asks one question over and over: "Given everything I've seen before, what word is most likely to come next?"

Think of it like the autocomplete on your phone's keyboard — except instead of being trained on your texts, it was trained on hundreds of billions of words from books, websites, Wikipedia, code, news articles, and more. The scale of that training is what makes the output feel like real writing.

This is called a Large Language Model (LLM). The "large" refers to both the amount of data used and the number of mathematical connections (called parameters) inside the model — GPT-4, for example, has an estimated 1.76 trillion parameters.

Token The basic unit an LLM works with. Not quite a word — more like a chunk of letters. "Unbelievable" might be split into "un", "believ", "able". AI reads and writes in tokens, not characters or full words.
Training Data The massive collection of text an LLM learned from before you ever talked to it. The patterns in that data shape everything the model knows — and everything it gets wrong.
Prompt The instructions or question you give to the AI. A better prompt almost always produces better output — which is why "prompt engineering" is now a real skill.
The Transformer Breakthrough

Before 2017, AI writing tools were clunky. They'd lose track of context after a few sentences. Then a team of Google researchers published a paper titled "Attention Is All You Need" — introducing the transformer architecture.

Transformers let AI pay "attention" to all the words in a sentence at once, not just the ones right next to each other. So when you write "The dog chased the ball until it burst," the AI can figure out that "it" refers to the ball, not the dog — because transformers look at the whole context together.

That single paper is the foundation of every major AI writing tool you've ever used: ChatGPT, Claude, Gemini, Copilot, and more.

Why This Matters For You

Knowing that AI predicts word-by-word helps you understand why it sometimes produces confident-sounding nonsense. It's not lying on purpose — it simply has no built-in fact-checker. It's completing a pattern, and sometimes that pattern leads somewhere wrong.

Is It Really "Writing"?

Writers, researchers, and philosophers debate this constantly. When an AI produces a paragraph, it hasn't experienced anything, felt anything, or chosen to express something meaningful. It ran math on probability distributions. But the output can still be useful, interesting, or beautiful — which raises the real question: does the process matter, or only the result?

For now, most educators, publishers, and companies say: AI output is a tool, not authorship. The human who prompted it, shaped it, and decided what to keep is the author. But that view is still evolving — and so are the rules around it.

💡 Student Tip

When you use an AI writing tool for school, always check its output for facts. AI can state things confidently that are completely made up. This is called a "hallucination" — a word we'll explore more in Lesson 3.

Quiz · Lesson 1

What Does AI Actually Do With Words?

3 questions — tap an answer to check it.
1. What is the primary way a language model generates text?
✓ Correct! AI writing is prediction at scale — the model has no understanding, just extremely sophisticated pattern-matching across billions of examples.
Not quite. Language models don't retrieve stored sentences — they generate new text one token at a time by predicting what comes next based on training patterns.
2. The 2017 paper "Attention Is All You Need" was important because it introduced the transformer architecture, which allowed AI to…
✓ Exactly right. Transformers use "attention" to weigh the relationship between all words simultaneously — which is why modern AI writing feels coherent across long passages.
That's not it. The key innovation was the attention mechanism — letting the model look at all words at once and figure out relationships between them, even words far apart in a sentence.
3. Why can an AI confidently write something that is completely false?
✓ Right. The model is optimized to produce fluent, plausible text — not accurate text. Confidence and correctness are two completely different things for an LLM.
Not quite. AI isn't trying to deceive anyone — it simply has no mechanism to verify facts. It generates whatever word sequence looks most plausible based on training patterns, even if that sequence is wrong.
Lab · Lesson 1

Exploring How AI Predicts Words

Chat with your AI lab assistant — 3 exchanges to complete the lab.

Your Mission

You've learned that AI writing tools predict what word comes next based on patterns. In this lab, you'll experiment with how prompts change predictions — and discuss what you notice with your AI lab assistant.

Try giving the AI assistant two versions of the same prompt (one vague, one specific) and ask it to explain what kind of writing it would produce. Pay attention to how much the output changes.

Try asking: "If I prompt an AI with just the word 'write' versus 'write a scary short story about a broken streetlight,' how different would the outputs be and why?"
AI Writing Lab
Lesson 1 — Prediction & Prompts
Hey! Welcome to the lab. 👋 I'm your writing-focused AI assistant for this module. We're going to explore how AI predicts words and why your prompts matter so much. What would you like to investigate first? You can ask me anything about how prompts change AI output, or tell me what's confusing about what you just read!
Lesson 2 · AI & Writing

How Writers Are Actually Using AI

Real authors, real newsrooms, real experiments — and what worked and what didn't.
When professional writers use AI as a tool, what do they gain — and what do they lose?

In the summer of 2023, the Writers Guild of America went on strike — one of the longest Hollywood strikes in decades. A central demand: clear rules about how studios could and couldn't use AI to generate scripts or revise writers' work. The studios wanted flexibility to use AI for first drafts; the writers argued this would erode wages and creative standards. After 148 days, a deal was reached: AI-generated material cannot be considered "literary material" under WGA contracts, and writers cannot be asked to rewrite AI output as if it were a human's draft. Real writers had drawn a line.

The Spectrum of Use

Not every writer who uses AI is "cheating" or "replacing themselves." In 2024, a survey by the Authors Guild found that roughly 20% of professional authors were experimenting with AI tools in some part of their process — but the ways they used it were very different.

Low-Stakes AI Use ✓

Brainstorming plot ideas, checking grammar, generating names for characters, creating outlines, summarising research, or getting unstuck during writer's block. Most writers who use AI this way say it speeds up the boring parts so they can focus on the creative parts.

High-Stakes AI Use ⚠

Generating entire article drafts, writing dialogue, producing poetry or literary prose, or submitting AI-written content under a human byline. This is where quality drops, readers notice, and ethical problems begin — including plagiarism and misrepresentation.

Newsrooms Draw Lines

In 2023, several major publications began publishing AI-generated articles without adequate disclosure. CNET quietly published 77 AI-written finance articles. When readers and staff discovered this, an audit found more than half contained factual errors — including wrong interest rate calculations and incorrect descriptions of how compound interest works. CNET issued corrections and paused the program.

The Associated Press took a different approach. They published a clear AI usage policy: AI can help with data processing, sports game summaries, and financial earnings reports — but every published story still requires a human editor to verify facts and take editorial responsibility. That framework has become a model for other news organizations.

The CNET Lesson

CNET's experience shows the core problem with AI in journalism: it sounds authoritative even when it's wrong. A reader can't tell the difference between a confidently-written correct fact and a confidently-written incorrect one. Human editors exist precisely to catch those errors.

Authors Who Experimented

In 2022, science fiction author Clarkesworld Magazine — one of the most prestigious short fiction venues in the genre — reported a massive surge in AI-generated story submissions. By early 2023 they had to temporarily close submissions entirely because the volume of AI-generated work overwhelmed human editors. Editor Neil Clarke described the AI submissions as "a flood of mediocre, soulless text."

Meanwhile, acclaimed author Robin Sloan publicly experimented with using language models as writing partners — feeding his own prose into a fine-tuned model and then using its suggestions as a "strange mirror" to reflect unexpected ideas back at him. He was careful to describe it as a tool that augmented his process, not one that replaced his voice.

The difference between those two cases — bulk generation vs. thoughtful integration — is the difference that most professional writers point to when they talk about AI use that's acceptable versus harmful.

💡 Student Tip — Your Own Writing

For school assignments, using AI to brainstorm ideas is usually okay (check your school's policy). Using AI to write your essay and submitting it as your own is dishonest — and also skips the learning your teacher is trying to give you. The goal of writing assignments isn't just the document — it's the thinking you do while writing it.

What Stays Human

Across all these real cases, one pattern emerges: AI is better at generating than at judging. It can produce 20 opening lines for a story, but it can't reliably tell you which one is best for your specific voice, your specific reader, your specific purpose. That judgment — the taste, the intention, the meaning behind the choices — stays stubbornly human.

Quiz · Lesson 2

How Writers Are Actually Using AI

3 questions — tap an answer to check it.
1. What did CNET discover after auditing its AI-generated finance articles in 2023?
✓ Correct. CNET's audit found factual errors in more than half of the 77 AI-written articles — including miscalculations about compound interest — which is exactly why human editorial oversight matters.
Not quite. The audit revealed significant factual errors — over half the articles had problems. This became one of the clearest real-world examples of why AI writing needs human verification.
2. What was the main outcome of the 2023 WGA (Writers Guild of America) strike regarding AI?
✓ Right. The WGA won key protections: AI output isn't "literary material" and writers can't be forced to treat AI drafts as if they were written by humans — protecting both wages and creative standards.
That's not what happened. The strike ended with a specific compromise: AI can't be treated as "literary material" under the contract, and writers can't be required to rewrite AI output as part of their job duties.
3. Based on the lesson, which use of AI in writing is most likely to cause problems?
✓ Correct. Submitting AI output as your own — without disclosure, verification, or meaningful human input — raises both quality and ethical problems. The CNET case and Clarkesworld flooding are both examples of this going wrong at scale.
Think about which use hides the AI's involvement and skips the human judgment step. The most problematic use is presenting AI text as wholly your own without verification or disclosure — which is what caused problems for CNET and overwhelmed Clarkesworld editors.
Lab · Lesson 2

AI as Writing Partner vs. Writing Replacement

Chat with your AI lab assistant — 3 exchanges to complete the lab.

Your Mission

In this lab, you'll explore the line between using AI as a tool versus letting it replace your thinking. Your AI assistant will help you brainstorm ideas for a short story — but your job is to notice when the ideas feel like yours and when they feel borrowed.

Pick a story idea you actually care about, share it with the assistant, and see what kinds of suggestions it offers. Then reflect: which suggestions feel useful, and which ones feel like they'd make the story less yours?

Try starting with: "I want to write a short story about [your idea]. What are three ways I could make the opening scene surprising? Then tell me which of those suggestions you think is most original."
AI Writing Lab
Lesson 2 — AI as Partner
Hello! Ready to explore the writing-partner idea. 📝 Share a story concept you're genuinely interested in — doesn't have to be polished — and we'll see what AI suggestions look like in practice. I'll offer ideas, but I'll also help you think about whether those ideas are actually useful for YOUR story or whether they'd pull it in a generic direction.
Lesson 3 · AI & Writing

Hallucinations, Bias & Why AI Gets Things Wrong

The confident wrong answer is more dangerous than obvious nonsense. Here's what's actually happening.
If AI sounds totally certain, how do you know when to trust it?

In May 2023, two lawyers at the New York firm of Levidow, Levidow & Oberman submitted a legal brief to federal court that cited six court cases as precedents. Judge P. Kevin Castel noticed that none of the cited cases appeared in any legal database. When the lawyers were questioned, they admitted they had used ChatGPT to research the brief — and the AI had invented all six cases, complete with realistic-sounding case names, docket numbers, and rulings. The lawyers were sanctioned. The judge called the submitted brief "gibberish" dressed up in legal language.

What Is a Hallucination?

In AI terminology, a hallucination is when a model generates information that sounds correct but is completely fabricated. The word comes from the idea that the AI is "seeing" things that aren't there — inventing facts, citations, names, or events that don't exist.

This happens because of how LLMs work: they're optimized to produce plausible text, not true text. When asked about a court case, the model generates the kind of text that typically describes a court case — judge name, docket number, ruling — whether or not that specific case exists.

Hallucination When an AI generates confident-sounding but completely fabricated information — fake citations, invented quotes, nonexistent events. A real and documented problem in every major LLM.
Confabulation A related term borrowed from psychology — when a brain (or AI model) fills in memory gaps with invented but believable details, without any intent to deceive.
Where Bias Comes From

AI writing tools inherit the biases present in their training data. Since most training data comes from the internet — which reflects human biases around gender, race, culture, and history — those biases get baked into the model's outputs.

One documented example: multiple studies in 2022 and 2023 found that when researchers prompted GPT models to generate stories with the instruction "write about a nurse," the AI defaulted to female pronouns far more often than male ones — reflecting the gender bias in its training data, not any factual distribution the model had been instructed to reproduce.

Similarly, researchers at Stanford found that models trained on English internet data tend to produce text that centers Western cultural assumptions — about family structure, success, relationships, and values — as if they were universal, when they're actually specific to particular cultures.

Bias Is Not a Bug — It's a Feature of the Data

AI companies didn't program in biases deliberately. The biases come from the text the model learned from — text written by humans who hold biases. This is why diverse training data, and human oversight of outputs, remains so important.

Strategies for Catching Errors

Because hallucinations are a built-in risk, smart users of AI writing tools develop verification habits. Here are the methods that researchers and journalists actually use:

  • 1Always verify any specific claim the AI makes — dates, names, statistics, events — in a second, reliable source. Never trust a number or a citation without checking.
  • 2Ask the AI: "Are you certain about this? What's your source?" If it can't cite a real, verifiable source, treat the claim as unverified.
  • 3Watch for "confident vagueness" — impressive-sounding general statements that dissolve when you try to verify the specifics. AI is especially good at producing these.
  • 4Look for bias by asking the AI to generate multiple perspectives on a topic. If it consistently favors one viewpoint without prompting, that may reflect training data bias.
  • 5Treat AI writing the way you'd treat a source that writes well but gets facts wrong sometimes — useful for ideas, not reliable for facts.
💡 Student Tip — Research & Reports

If you use AI to help research a report or essay, never paste an AI-generated list of sources into your bibliography. Verify every single source exists. The fake-cases lawsuit is an extreme example — but students have submitted AI-invented sources in papers too. Your grade and your credibility are worth the extra check.

Quiz · Lesson 3

Hallucinations, Bias & Why AI Gets Things Wrong

3 questions — tap an answer to check it.
1. In the 2023 legal case from the lesson, what did ChatGPT fabricate?
✓ Correct. ChatGPT invented six complete, realistic-sounding legal cases — none of which existed. The lawyers trusted the AI's confident output without verifying it, and were sanctioned by the court.
Not quite. The AI invented six entirely fictional court cases, complete with realistic details like docket numbers. This is a classic hallucination — confident, detailed, and completely fabricated.
2. Why do AI language models produce hallucinations?
✓ Exactly. LLMs predict the most plausible next token — they have no separate fact-checking layer. Plausible and true are very different things, which is why confident-sounding fabrications are possible.
Think about how LLMs work — they predict what sounds right. They're not programmed to deceive, and the problem isn't about memory. The core issue is that plausibility and truth are different things, and LLMs only optimise for plausibility.
3. What is the best description of where AI bias comes from?
✓ Right. Bias isn't designed in — it's inherited from training data written by humans who hold biases. This is why diverse data and human oversight of outputs matters so much.
Bias in AI isn't intentional programming — it comes from the training data. Since humans wrote that data, and humans have biases, the model absorbs those patterns. Modern LLMs still show significant bias in documented studies.
Lab · Lesson 3

Spot the Hallucination

Chat with your AI lab assistant — 3 exchanges to complete the lab.

Your Mission

Your goal in this lab is to try to catch the AI making something up. Ask your lab assistant a factual question about a specific topic you already know well — a book you've read, a sport you follow, a historical event from class — and see if any of the details it gives you are wrong.

If you find an error, tell the assistant what's wrong. If it seems right, ask it to get more specific — hallucinations often appear when you push for details.

Try asking: "Tell me five specific facts about [a topic you know well]. Be as detailed as possible." Then check each fact carefully and report back what you found.
AI Writing Lab
Lesson 3 — Hallucination Detection
Let's do some fact-checking! 🔍 Pick a topic you actually know well — something you've studied, a hobby, a book, a historical period — and ask me to give you specific facts about it. Your job is to be the fact-checker. I might get things wrong (I'm an AI — I hallucinate sometimes!), and catching those errors is exactly the point of this lab. What topic should we test?
Lesson 4 · AI & Writing

Prompt Craft — Getting Better Writing from AI

The quality of your output is mostly determined before the AI types a single word.
What separates a prompt that produces something useful from one that produces forgettable mush?

By 2023, "prompt engineer" had become a real job title at major tech companies, with salaries reaching $300,000 a year at some AI firms. Anthropic, Google, and OpenAI each published internal research on what makes prompts effective. Their shared finding was counterintuitive: longer, more specific prompts almost always produce better results than short, vague ones — even though users naturally assume a "smarter" AI should understand what they want from less. The research confirmed that clarity, context, and constraints are the three pillars of effective prompting.

The Three Pillars of a Good Prompt

You don't need to be a professional to write effective prompts. You just need to understand what information the model actually needs to do a good job.

  • 1Clarity — what you actually want. "Write something interesting" is almost useless. "Write a 200-word opening paragraph for a mystery story set in a rainy city" gives the model something to work with. Be specific about format, length, and topic.
  • 2Context — who, where, and why. "Explain photosynthesis" produces a textbook answer. "Explain photosynthesis to a 10-year-old using a pizza analogy" produces something genuinely useful. Context shapes tone, vocabulary, and approach.
  • 3Constraints — what you don't want. "Write a product description without using the word 'innovative'" or "Keep this under 100 words" are constraints that prevent the most common AI clichés. Constraints are surprisingly powerful.
Prompting Techniques That Actually Work
Few-Shot Prompting

Give the AI an example of what you want before asking it to do it. "Here's an example of the writing style I like: [example]. Now write something similar about [topic]." This anchors the output to a specific tone and style.

Role Assignment

"You are an editor at a teen fiction magazine reviewing a short story for pacing problems." Giving the AI a specific role with a specific audience changes its output dramatically — it shifts both vocabulary and focus.

Chain of Thought

"First, outline the three main arguments. Then, write a draft of each one. Finally, write a conclusion that ties them together." Breaking a task into steps produces more coherent output than asking for everything at once.

Critique Mode

"Here is a paragraph I wrote. Tell me three things that are working well and two things that could be stronger." Using AI to critique your own writing — rather than write for you — keeps you in control while getting useful feedback.

Common Prompt Mistakes

Being too vague: "Write me a story" gives the AI almost nothing to work with. It will produce something generic. The vaguer your prompt, the more average the output.

Assuming the AI knows your context: The AI doesn't know you, your class, your teacher's expectations, your previous drafts, or your reading level unless you tell it. Treat it like a very knowledgeable stranger who knows nothing about your specific situation.

Accepting the first output: Treat the first response as a draft, not a final product. Prompt engineers often spend more time refining and re-prompting than they do on the initial ask. Ask for revisions. Specify what's wrong. Iterate.

The Real Skill

The writers and researchers who use AI most effectively aren't the ones who found a magic prompt. They're the ones who learned to have a conversation with the tool — asking, critiquing, redirecting, and shaping the output through multiple exchanges. That's a skill that looks more like editing than typing.

💡 Student Tip — Prompting for Learning

Instead of asking AI to write your essay, try asking it to be your Socratic teacher: "I'm writing about the causes of World War I. Ask me questions that help me figure out my argument." This keeps you thinking, uses the AI's knowledge as a scaffold, and produces writing that's genuinely yours.

Quiz · Lesson 4

Prompt Craft

3 questions — tap an answer to check it.
1. What do researchers at Anthropic, Google, and OpenAI consistently find about prompt length?
✓ Correct. Internal research from all three major AI labs confirms that clarity, context, and specificity in prompts consistently outperform short, vague instructions — even though users often assume smarter AI needs less direction.
The research finding is counterintuitive but consistent: longer, more specific prompts produce better output. The model has more to work with when you give it more context and constraints.
2. What is "few-shot prompting"?
✓ Right. Few-shot prompting means showing the model an example (or a few examples) of exactly what you want before your actual request. It anchors the tone, style, and structure of the output.
Few-shot prompting is about giving examples first. You show the AI what good output looks like before asking it to produce something — "here's an example of the style I want, now do something similar about X."
3. Which approach to AI writing keeps YOU most in control of your own thinking?
✓ Exactly. Critique Mode keeps you as the author — you do the thinking and drafting, and use AI's feedback to improve your own work. This is how many professional writers use AI productively without losing their voice.
Think about which option keeps your thinking central. If AI writes the draft or the outline, it shapes your ideas before you've formed them. Critique Mode — where AI responds to YOUR writing — keeps you in the driver's seat.
Lab · Lesson 4

Prompt Engineering Challenge

Chat with your AI lab assistant — 3 exchanges to complete the lab.

Your Mission

In this lab, you'll practice the three pillars: clarity, context, and constraints. You'll write the same request three times — once badly, once okay, once really well — and compare the results with your lab assistant.

Pick any writing task (a poem, an opening paragraph, a product review, an essay introduction). Write a vague version, a medium version, and a highly specific version. Ask the assistant to evaluate which prompt is likely to produce the best output and why.

Start with: "I'm going to give you three versions of a prompt for [writing task]. Tell me which one is the best prompt and specifically what makes it work better than the others." Then share all three versions.
AI Writing Lab
Lesson 4 — Prompt Engineering
Welcome to the prompt engineering lab! ✍️ This is where craft meets strategy. Pick any writing task you want — a story opening, a poem, an argumentative paragraph — and write three versions of a prompt for it: one weak, one medium, one strong. Share them all with me and I'll analyse what makes each one more or less effective. Ready when you are!
Module Test · AI & Writing

Module 2 — Final Assessment

15 questions · Pass score: 80% (12/15) · Tap an answer to lock it in.
1. What does an LLM (Large Language Model) actually do when it generates text?
✓ Correct — LLMs are sophisticated next-token prediction engines trained on vast datasets.
LLMs predict the next token based on training patterns — they don't retrieve stored sentences or copy from the internet.
2. What was the key innovation of the 2017 paper "Attention Is All You Need"?
✓ Correct — transformers use attention mechanisms to process all words in context at once, enabling coherent longer-form writing.
The key innovation was the transformer architecture and its attention mechanism — not internet access or emotional understanding.
3. In AI terminology, a "token" is best described as:
✓ Right — tokens are sub-word chunks that LLMs read and write in. A word like "unbelievable" might split into three tokens.
A token is a chunk of text — often a word fragment — that the model processes as a unit. Not a full sentence and not a single character.
4. What did CNET discover when it audited its 77 AI-generated finance articles in 2023?
✓ Correct — over half of CNET's AI-written articles contained errors, demonstrating the need for human editorial oversight.
The audit revealed factual errors in more than half the articles — not plagiarism, not accuracy. This is the hallucination and confidence problem in action.
5. The 2023 WGA strike resulted in what specific AI-related protection for writers?
✓ Correct — the WGA won protections preventing AI output from being classified as literary material and preventing writers from being forced to work from AI drafts.
The key outcome was that AI output can't be "literary material" under the contract and writers can't be required to revise AI drafts — not a total ban.
6. What happened when lawyers submitted a brief researched using ChatGPT to federal court in 2023?
✓ Right — this is one of the most documented hallucination cases in AI history. The AI invented six complete, realistic-sounding legal cases that didn't exist.
ChatGPT hallucinated six entirely fictional court cases, complete with case names and docket numbers. The lawyers were sanctioned for submitting them without verification.
7. What is an AI "hallucination"?
✓ Correct — hallucinations are fabricated-but-plausible outputs: invented facts, citations, quotes, or events that the AI presents confidently.
A hallucination is when AI generates fabricated information that sounds completely real and confident — not a glitch, not a refusal.
8. Where do bias patterns in AI writing tools originate?
✓ Right — AI bias is inherited from training data. Human text carries human biases, and the model learns those patterns along with everything else.
Bias in AI isn't programmed intentionally — it's absorbed from training data written by humans who hold real biases. The model learns those patterns.
9. The Clarkesworld Magazine experience in 2023 illustrated what problem?
✓ Correct — bulk AI generation flooded the submission queue with what editor Neil Clarke called "soulless" text, forcing a temporary closure.
The problem was volume — bulk AI-generated submissions overwhelmed human editors, forcing a temporary shutdown. A clear example of misuse vs. thoughtful integration.
10. What are the three pillars of an effective AI prompt, according to the lesson?
✓ Correct — clarity (what you want), context (who/where/why), and constraints (what you don't want) are the three pillars of effective prompting.
The three pillars are clarity, context, and constraints — telling the AI what you want, giving it relevant background, and specifying what to avoid.
11. What is "few-shot prompting"?
✓ Right — few-shot prompting shows the AI what good output looks like first, anchoring tone, style, and structure before the main request.
Few-shot means showing examples first. "Here's the style I want — now do something similar about X" is few-shot prompting.
12. According to researchers at major AI labs, what do users often assume incorrectly about AI prompts?
✓ Correct — the counterintuitive finding is that more specific, longer prompts produce better results, even though users assume smarter AI needs less guidance.
The key incorrect assumption is that better AI needs less instruction. Research consistently shows the opposite — more specific prompts produce better output.
13. Which of these is the best example of "Critique Mode" for a student using AI on an essay?
✓ Exactly right — Critique Mode keeps you as the author. You do the thinking and drafting; AI responds to your work rather than replacing it.
Critique Mode means you write first, then ask AI to respond to YOUR writing. Sharing your draft paragraph for feedback is Critique Mode — it keeps your thinking central.
14. The Associated Press's AI policy is considered a model for news organizations because it:
✓ Right — the AP model allows specific AI uses while keeping human editorial responsibility mandatory. That balance is what makes it influential.
The AP found a middle ground: use AI for defined low-risk tasks, but always require human editorial verification and responsibility before publication.
15. What is the best overall description of what stays "stubbornly human" even as AI writing tools improve?
✓ Correct — AI can generate, but it cannot judge. The taste, intention, and meaningful choices that make writing worth reading remain a human function.
What stays human is judgment — the taste to know which of twenty AI options is right for your specific purpose, audience, and meaning. AI generates; humans judge.