Loading page…
AI-written books on Amazon riding platform traffic: zero inventory, zero English barrier — pick niches by rank, cold-start on the first 10 reviews
"Best-sellers usually aren't the most useful products — they're the ones that best feed the imagination." An $8 blank notebook titled 'Words My Younger Self Needs to Write Down for When I'm Old and Remember Nothing' — what people buy is the emotion of the instant they read the title.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
[AI] This project comes from a Douyin video: the creator says he watched an anonymous "Asian guy's (CN)" YouTube channel three times, then broke it down — claiming the guy made $2.2 million over three years writing books with AI. The protagonist is anonymous, no channel name, no revenue screenshots — the $2.2M is purely the creator's retelling. It made the site under the new criteria: Amazon KDP is a long-validated passive-income arena, AI drives writing costs to near zero, the barrier is minimal, and the platform brings its own traffic — my judgment is that "AI book writing" as a direction has potential, not that this specific number is credible. Realistic earnings are more likely a long tail of a few hundred to a few thousand dollars a month.
Amazon supplies the customers; the work is deciding which search result you want to be. Americans spend about $83 a year each on books and already pay for digital goods without hesitating, so the whole acquisition problem collapses into on-site ranking. The method is research before writing: ask ChatGPT for 50 candidate topics, pick the one you actually have something to say about, then check it in Amazon search with the Book Bolt plugin, which shows the rank and sales of every book under a keyword. The threshold used is a rank inside the top 80,000, worth roughly $500 a month; if five or more books under one topic clear it, the topic has demand and room. After that the listing does the selling, because the title and cover are the only surface a buyer sees, as with the $8 blank notebook sold as "Words My Younger Self Needs to Write Down for When I'm Old and Remember Nothing." There is no ad budget at an $8 price point, and nothing else to spend it on.
The first dollar is a royalty on an $8 impulse purchase made by someone reacting to a title. KDP charges no registration and no monthly fee, so the distance from zero to a first sale is a manuscript and a cover. The manuscript itself takes under an hour: ask the AI what eight steps carry a reader from problem A to outcome B, treat those eight steps as eight chapters, and write one chapter per conversation rather than generating everything at once. The video is blunt that even an unedited pure-AI draft uploaded as-is can earn something like $100. Scale comes from rank, not price, and the benchmarks quoted are per-title daily numbers: about $200 a day for that $8 blank notebook, about $1,500 a day for a decluttering title sold as "Nobody Wants Your Junk." What the buyer is paying for is the feeling at the moment of reading the cover, which is why roughly 90 of every 100 buyers never finish the first chapter and still do not ask for a refund.
Revenue unverified, but KDP is a decade-proven passive-income arena, and AI drives the cost of "writing" to near zero — exactly what you want: "low barrier + the platform brings the customers". Best as a branch line of your pipeline: use AI to batch-produce "low-content books" (notebooks/puzzles/planners) to test the waters — nearly zero copyright and quality risk — and only move to proper content books once the niche-picking logic works (Book Bolt, top-80K rank). Don't touch review manipulation or cover copying.
Paste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
Data credibility: ⚠️ Unverified: the creator relays an anonymous "Asian guy" case with no proof of the $2.2M; listed because the AI judges KDP a mature, monetizable arena. All revenue figures are founder-reported or estimated by the video creator, unaudited.
The protagonist had been publishing on KDP for six years, and the money only arrived in the last two, when the writing step stopped being the bottleneck. The arithmetic he is quoted on is the whole thesis: one book a month is about $500, so the question becomes what 100 books are worth. Each title is a permanent asset that keeps drawing search traffic, and a catalog compounds in a way a single book cannot. Two smaller loops sit inside that. Reviews are the first: nobody buys a product with zero reviews, so listing free until at least ten reviews land, then switching to paid, converts the herd effect into rank, and rank brings more reviews. Note that Amazon's terms treat incentivized or manipulated reviews as a violation, so this tactic runs close to the line. Format is the second: one manuscript becomes an ebook, an audiobook and a print-on-demand paperback, with printing, packaging, logistics and returns all handled by Amazon, so a book that works earns three ways without extra production.
The stretch here lasted four years before it ended. The protagonist started on KDP six years back and did not make real money until AI collapsed the cost of a draft, which is the honest shape of this business: the early years are cheap and quiet rather than dramatic. The signal to watch is not sales but rank, since a position inside the top 80,000 stands in for roughly $500 a month, and the first ten reviews are the verdict on whether the book actually solved anything. The adjustment that gets made is narrowing, not rewriting. In the walkthrough, "overcoming procrastination" was judged too broad to differentiate and became "how people who don't work a regular job beat procrastination." At $8 and near-zero marginal cost, the correct response to a dead title is another title into a better-researched keyword, while the ones already ranking keep earning. As the video puts it, AI lets you lose in a great many races, and you only need to win once.
[Video] (1) Niche picking (most important): use Amazon's on-site search + the Book Bolt plugin to check the rank and sales of every book under a keyword; rule of thumb, "a rank inside the top 80,000 ≈ roughly $500/mo" — if 5+ books under one topic crack the top 80K, the topic is doable. "Best-sellers usually aren't the most useful products — they're the ones that best feed the imagination": an $8 blank notebook titled 'Words My Younger Self Needs to Write Down for When I'm Old and Remember Nothing'; a decluttering book called 'Nobody Wants Your Junk'. (2) Writing: ChatGPT picks the topic → the A-to-B bridge method asks the AI "what 8 steps take someone from bad state A to goal B" — those 8 steps become the 8-chapter outline → write one chapter per conversation (don't generate it all at once) → adding personal stories to a pure AI draft lifts the "human feel" from 60 to 85 points; non-English speakers write in Chinese → AI translates → a second AI cross-checks. (3) Cold start: list it free first, bank at least 10 reviews, then flip to paid — exploiting the herd effect (nobody buys at zero reviews; once good reviews exist, conversion soars into a positive loop).
[Video] Categories: ebooks + audiobooks + POD print books (printing, packaging, logistics, and after-sales all handled by Amazon), plus "low-content books" — puzzle books, crosswords, blank notebooks, planners — nearly pure templates. Tools: ChatGPT (writing), Amazon KDP (publishing), Book Bolt (niche research via ranks and sales), design AI (Lovart-type, for covers/puzzles/crosswords). [AI] Low-content books are your safest entry point: almost no text, no copyright risk, no quality blow-ups — it's purely niche picking and cover emotion — ideal for testing the waters with assembly-line batch output.
[AI] (1) The $2.2M can't be traced — don't set expectations on that number. (2) Compliance red lines: Amazon KDP has disclosure requirements for pure-AI content and policies against imitating best-seller titles/covers, and review manipulation (the video's "bank 10 free reviews") explicitly violates Amazon's terms — any of these can get your account/books taken down; in practice, stay away from review manipulation and cover copying. (3) The arena is grinding: AI lets anyone mass-publish, low-quality books are flooding in, and niche picking plus differentiation decide who wins. Your move: first use AI to batch-produce compliant low-content books (notebooks/puzzles/planners) to get the niche-picking logic working, then consider proper content books.
That was one of the 50 breakdowns open to everyone. The Unicorn Club opens the other 1,285 at the same depth — plus pain-point filters and 10 build specs a week.
Join the Unicorn Club — $5/moCancel anytime, in one click