Chapter 13 — Exercises

Work these in order; they climb from recall to real judgment. You will need a web browser and, ideally, a site you can inspect. Everything here uses free tools. Items marked with a dagger (†) have full worked solutions in Appendix — Answers to Selected Exercises; odd-numbered items are worked there too. Attempt every item before you look.

The spirit of the chapter, kept close: the question is never "did a machine write this?" It is "is this genuinely helpful, original, and accurate — and is there a real, accountable human behind it?"

A. Recall and vocabulary

Exercise 13.1 — Define, in your own words, three terms this chapter owns: AI-generated content, content at scale, and scaled content abuse. Explain why the first two are neutral and only the third names a policy violation.

Exercise 13.2 † — State Google's documented position on AI content (paraphrase the thesis in the epigraph), then explain the "Who, How, and Why" framework in one sentence each. Which of the three questions is "the tell," and why?

Exercise 13.3 — Define information gain. Explain why a large language model is structurally weak at producing it, referring to what a model actually is.

Exercise 13.4 — Define human-in-the-loop and editorial judgment. In the workflow from §13.5, which step is the "expensive" part of the job, and why can AI not perform it?

B. Applied reasoning

Exercise 13.5 — Rewrite this sentence so it is accurate by the chapter's standards: "AI-generated content is against Google's guidelines and will get you penalized."

Exercise 13.6 † — Distinguish content at scale from scaled content abuse using two examples that involve the same number of pages (say, 5,000). Make the only difference the thing that actually matters.

Exercise 13.7 — The chapter says AI "changed the cost of playing the game badly," not "changed the game." Explain what that means, and connect it to the book's first theme (SEO is not a trick).

Exercise 13.8 — A colleague insists on running every draft through an AI detector before publishing. Give two separate reasons this is misguided — one about the tools, one about what Google actually does.

Exercise 13.9 — Explain why, even if a perfectly accurate AI detector existed, it would not change how a responsible practitioner produces content.

Exercise 13.10 † — The chapter argues that as generic content became abundant, its value as a differentiator "collapsed," while the value of genuine experience rose. Explain the economic logic in your own words, and say what a creator should conclude from it.

C. Diagnose the problem

Exercise 13.11 — A site published 800 AI-generated articles in two months, each targeting a keyword scraped from a competitor; six months later its traffic collapsed after a core update. Diagnose: which policy did it violate, was it "penalized for using AI," and is there a fix that preserves the model?

Exercise 13.12 † — Here is an AI-generated draft paragraph for Rivertown's "furnace won't ignite" guide (constructed teaching example): "If your furnace won't ignite, there are several possible causes. Check the thermostat settings and make sure the furnace has power. The igniter or pilot light may be the issue. It is often best to consult a professional if the problem persists." Identify three things missing that only a licensed technician could add, and name the E-E-A-T component each one supplies.

Exercise 13.13 — A genuine small publisher with real, first-hand expertise lost traffic in the same update as an obvious spam farm. Does this prove Google "can't tell the difference," so quality is pointless? Give the honest, two-sided reading the chapter demands.

Exercise 13.14 — A marketer describes their workflow: "We prompt the model, a junior editor skims each draft for typos, and we publish forty articles a week." Name the two workflow steps they have effectively removed, and predict the likely outcome.

D. On your own site (use a free tool)

Exercise 13.15 † — Take a page you personally wrote by hand (ideally from before generative AI was common) and paste a few paragraphs into two or three free "AI content detector" tools. Record each verdict. Write one sentence on what the disagreement — and any false "AI" flag on your own writing — tells you.

Exercise 13.16 — Open your most important page and interrogate it with the "Who, How, Why" questions: Who is accountable for this? How was it made, and does the page help a reader understand that? Why does it exist — to help someone, or to catch clicks? Write down every place the page fails to make the author or the experience legible.

Exercise 13.17 — Search a query your site targets and note whether an AI Overview appears. Write down what the AI Overview covers and one follow-up question a real searcher would still have that the summary leaves unanswered — the gap your content could own.

E. Read the report / interpret / rewrite

Exercise 13.18 † — Interpret this report.

FIGURE E13.1 — "Two content sites after a quality update"          [labeled composite — real patterns]
  SITE X: 3,500 pages generated by prompting a model against a keyword export; no named authors; every page
          restates the top-ranking consensus; no original data, photos, or first-hand accounts. Down ~90%.
  SITE Y: a 400-page site by one genuine practitioner — real testing, original photos, named author — but
          recently bulked up with 150 thin AI "explainer" posts to "cover more terms." Down ~45%.

For each site, say whether the demotion was "the system working as designed" or "the harder case," and state the single most sensible move.

Exercise 13.19 — Rewrite. Here is a generic AI-drafted opening for a "how to relight a gas water-heater pilot light" page (constructed): "Relighting a pilot light is a simple process. Locate the gas control valve, turn it to pilot, hold down the reset button, and light the pilot. If it won't stay lit, you may need to call a professional." Revise it to add first-hand experience and information gain. Invent plausible practitioner detail, and label your additions as illustrative.

Exercise 13.20 — Research/analysis. Choose an informational query in your field. Read the top result and any AI Overview. List what original information — data, first-hand testing, a specific real example — is absent from all of it. That list is your information-gain opportunity; note which item you could actually supply.

F. Myths, ethics, and judgment

Exercise 13.21 † — Spot the myth. "Running content through an AI detector verifies it is safe from Google penalties." Identify the two false assumptions bundled in that sentence and correct each.

Exercise 13.22 — Spot the myth. "Using AI to draft anything is cheating, and AI content is inherently unethical." Explain why this overcorrects, and state where the ethics actually live (§13.1's "why," §13.7).

Exercise 13.23 — Ethics judgment. A site prompts a model into confident, authoritative-sounding medical-advice articles and publishes them under a fabricated "Dr." byline. Name every guardrail this violates, and cite the chapters that own each concern.

Exercise 13.24 † — You publish an AI-assisted article and a reader later finds it contains a fabricated statistic you didn't catch. Whose error is it now, and what does the chapter say your responsibility was before you hit publish?

Exercise 13.25 — Explain the "don't flood the commons" argument in your own words: why does the chapter call mass-publishing thin AI content "defecting against the ecosystem your own business depends on"?

G. The Strategy File (Rivertown Home Services)

Exercise 13.26 † — Write a one-paragraph brief (step 1 of the §13.5 workflow) for a Rivertown AI-assisted guide on "why is my AC freezing up." Specify the intent, and — most important — the real experience Tony Delgado or a lead technician must supply that a model cannot know.

Exercise 13.27 — Rivertown could generate 300 blog posts this week with AI. Explain, using Chapter 12's content-audit logic, why doing so would rebuild the exact problem the blog audit just solved.

Exercise 13.28 — Name the Rivertown content topics on which a licensed human must own the content outright — not merely review an AI draft — and explain why, referencing YMYL (Chapters 5 and 35).

Exercise 13.29 † — Extend the Strategy File: propose how Rivertown makes the technician's review visible on the page (the E-E-A-T legibility step). List the specific on-page elements, and say what each one demonstrates.

Exercise 13.30 — Rivertown's owner asks: "If AI drafts the content, can we cut our content budget?" Give the honest answer using the chapter's cheap-part / expensive-part distinction.

Where these go next. Keep your answers to 13.16, 13.17, 13.26, and 13.29 — they extend your running Strategy File (Appendix C's worksheet). The workflow you sketch here is the same one you will formalize when you assemble the full content plan for your own site alongside Rivertown's in Chapter 40.