By Han Jong-woo | April 2026
SummitSelect.org | AI Tools | eBook Publishing | Writing & Income
The Bottom Line — Read This First
I published four eBooks on Amazon in the past two years. At 71.
Not because I have a publishing background. Not because I’m a fast writer. Not because I had months of free time to dedicate to each one.
Because I figured out how to use AI as a collaborator in a way that dramatically accelerated the production process without compromising the one thing that makes a book worth reading: the genuine human knowledge, experience, and perspective that only the author can provide.
I want to be precise about this from the start, because there’s a version of “creating eBooks with AI” that produces generic, forgettable content that nobody benefits from reading. And there’s a version that produces genuinely useful books, faster than traditional authorship allows, because AI handles the parts of book creation that don’t require human experience — and you handle the parts that do.
This article is about the second version. Everything I know about it came from actually doing it — from the mistakes of my first book, the refinements of my second, and the more efficient process I’d developed by the time I was working on my fourth.
If you’re thinking about writing an eBook — whether as an income stream, as a way to share expertise you’ve accumulated, or both — this is the honest, experience-based guide I wish I’d had at the beginning.
Introduction: Why I Started Writing eBooks in My Late 60s
The backstory matters here, because it explains both why I started and why AI was essential to making it work at my age.
I’d been writing for SummitSelect.org for about two years — blog posts, guides, articles on AI tools for older adults. Over time, the topics I was writing about most — how seniors could use AI in daily life, how to generate income in retirement, how to navigate healthcare with AI assistance — were too substantial for articles. They needed more depth, more structure, more room to breathe than a blog post allows.
I also started hearing from readers who wanted something they could sit with, something organized comprehensively rather than topic by topic across dozens of separate articles.
The natural response was a book. But the traditional timeline for writing a nonfiction book — a year or more of research, writing, and revision — felt incompatible with the pace at which both AI and the needs of my readership were evolving.
I needed a way to produce something book-length and book-quality in weeks rather than months. AI made that possible without cutting corners on the things that actually matter.
Step One: The Idea Validation You Can’t Skip
The single most common mistake I see people make when they talk about using AI to write eBooks is starting with AI.
They open ChatGPT, type “write me an eBook about [topic],” and receive something that is technically book-shaped and essentially worthless — because it contains no specific knowledge, no genuine perspective, no real experience, and nothing that a potential reader couldn’t find through a five-minute Google search.
The right starting point is the opposite of AI. It’s human.
Before I touch any AI tool for a new book, I spend several days answering a set of questions in my own handwriting.
What do I actually know about this topic that most people don’t? Not what I’ve read. What I’ve personally experienced, figured out, and tested. The specific knowledge that comes from having been in the situation.
Who is this book for, specifically? Not “people interested in AI” — that’s millions of people with wildly different needs. “Adults over 65 who have heard about AI but have never used it, who are intimidated by technology, and who want to know if any of it is actually relevant to their daily life.” That specific.
What does this person most need to know that they don’t know? What’s the gap between where they are and where they need to be? What’s standing between them and a better situation?
What will they be able to do after reading this that they couldn’t do before? The clearest test of whether a nonfiction book has a real purpose.
If I can answer all four of those questions specifically and honestly, I have a book worth writing. If I can’t, I spend more time on this step until I can.
This step takes two to four days for me. It’s not efficient. It’s essential. No AI tool can do it for you because the answers come from your specific knowledge and your specific understanding of your specific audience.
Step Two: The Outline — Where AI Earns Its Place
Once I have clear answers to those four questions, AI becomes genuinely useful. And the first place it earns its place is in building the outline.
Here’s my specific process.
I write my own rough chapter list first. Not detailed — just the major areas I know need to be covered, in the order that makes sense to me based on my knowledge of the subject. For my book on AI and health, this was: understanding what AI actually is, how to use it to understand medical information, how to prepare for appointments, medication management, navigating insurance, and safety considerations.
Then I take that rough list to ChatGPT or Claude and give it the full context: my specific audience, my book’s purpose, the gap I’m filling, and my rough chapter list. I ask for a detailed outline — full chapter structure with H2 and H3 subheadings — that would serve this audience and accomplish this purpose.
What comes back is usually 60 to 70 percent right. The structure is often better than my initial rough list. The chapter sequence is frequently more logical. But there are always things it misses — nuances specific to my audience that require the specific knowledge I brought to step one.
I spend about two hours reviewing, reorganizing, adding, removing, and refining the AI-generated outline until it accurately reflects the book I actually want to write.
That refined outline becomes the architectural document the whole book is built from. Every chapter, every subheading, every section has a defined purpose before I write a single word of content.
Why a Good Outline Matters More Than People Think
Writing without a solid outline is like building a house without blueprints. You might produce something that stands up, but you’ll spend enormous time on structural problems that good planning would have prevented.
With a solid outline, every writing session has a specific, bounded task. Not “write the chapter on medication management” — “write the section on how to use AI to check for drug interactions, for an audience of adults over 65 who are taking multiple medications and have never used AI tools before.” That specificity makes the writing session manageable and makes AI assistance within it far more useful.
Step Three: Writing the Content — The Human-AI Division of Labor
This is the step where the process gets most nuanced, and where I want to be most precise.
For each section of the book, I follow a consistent sequence.
First, I write the core. Before AI touches the section, I write — in my own words, from my own experience — the essential knowledge that section needs to convey. Not polished prose. Notes, really. What do I know about this? What specific experience do I have? What do I want the reader to understand that they couldn’t get from a generic source?
For a section on preparing for medical appointments, my core notes looked like: “I’ve been doing this for two years. The first time I prepared using AI was before my cardiologist appointment in early 2024. I had the AI help me organize my symptoms and generate questions. My doctor said it was the most organized he’d seen me. That one experience convinced me this was worth teaching. The key insight: most patients go to appointments reactive, not prepared. AI flips that.”
That note — rough, personal, specific — is the genuine content of the section.
Then I use AI to develop it. I paste my notes into the AI tool along with the section’s purpose and audience, and I ask it to develop the content into a full section — expanding my core ideas into clear, comprehensive explanation while maintaining the practical, experience-based tone.
Then I rewrite it. The AI draft is a starting point, not a final product. I go through every paragraph. Anything that sounds generic, I make specific. Anything that sounds like AI-written content — slightly too polished, too balanced, without the rough edges of genuine opinion — I rough up. I add the personal examples that AI can’t know. I cut anything that’s padding.
The final section is genuinely mine — informed, expanded, and structured by AI assistance, but authored in my voice with my knowledge and my specific examples.
This process produces one to two sections per hour. For a 30,000-word book (approximately 100 to 120 pages), that’s 30 to 40 sections, meaning 15 to 25 hours of drafting work. The actual total time, including outlining, research, and revision, runs four to six weeks at a pace of two to three hours per day.
That’s a book in a month to a month and a half. Without AI, the same book would have taken me six months at minimum.
Step Four: The Editing Process — More Human Than People Expect
When people think about AI and editing, they assume AI handles it. That’s backwards.
AI is useful for specific, bounded editing tasks. It is not a substitute for the human editorial judgment that determines whether a book is actually good.
Here’s how I divide the editing work.
AI handles: Proofreading for spelling and grammar. Identifying places where the same point is made multiple times unnecessarily. Flagging passive voice constructions. Checking whether the tone is consistent across sections.
I ask specifically: “Read this chapter and tell me the three places where a reader is most likely to lose engagement, get confused, or feel like the content is repetitive. Be specific.” The response gives me a specific editing agenda.
I handle: Determining whether the content actually serves the reader’s real need. Evaluating whether the examples are the right examples. Deciding what to cut — which is the hardest and most important editorial decision. Reading every sentence aloud to check whether it sounds like me or like a content machine.
The cutting, in particular, is something AI consistently gets wrong when you ask it to edit at the level of what to remove. It tends to conserve — to find reasons to keep things — because its training doesn’t include the editorial judgment that says “this section isn’t earning its place.” That judgment requires a human who knows what the book is trying to accomplish and can evaluate whether every part of it is contributing to that goal.
My editing process typically takes as long as the drafting — four to six weeks — because I do multiple passes, with time between them to read with fresh eyes.
Step Five: Formatting and Cover Design
This is where practical AI assistance meets practical tool knowledge.
Formatting for Amazon KDP
Amazon Kindle Direct Publishing (KDP) accepts Word documents and converts them to Kindle format. I write everything in Microsoft Word and format for KDP using their provided style guide.
AI is useful here for the specific task of generating the front matter — the title page, table of contents, copyright page, and author biography — in the correct format. I provide the content; AI helps me structure it correctly.
The chapter headings, internal formatting, and spacing need to follow specific KDP requirements. I’ve created a template from my first book that I reuse — meaning formatting for books two, three, and four took less than two hours each.
Cover Design
I cannot overstate how much a cover matters for eBook sales on Amazon. The cover is the primary visual signal that tells a potential reader whether this book is worth their time and money.
I use Canva for cover design. Canva’s AI features — particularly the ability to generate background images and adjust design elements through natural language — have made cover design accessible to someone without a design background.
But I’ve also learned that the most important cover decision — the specific visual concept that communicates what the book is — requires human judgment about the audience. For AI Made Simple for Seniors, I tested three different cover concepts with a small group of readers from my target demographic before committing to the final design. The version they responded to was not the one I would have chosen without their input.
Step Six: Writing Copy That Sells
Amazon is a search engine for books. The description, the subtitle, the keywords, and the categories you choose determine whether your book gets found.
This is another area where AI is genuinely useful — and where most first-time eBook authors significantly underinvest.
The Book Description
The Amazon book description is your primary sales copy. It needs to answer three questions in the first two sentences: who is this for, what problem does it solve, and what will they be able to do after reading it.
I draft my own version first, then use AI to generate five alternative versions. I rarely use any of them verbatim. But the exercise of generating alternatives consistently produces specific phrases or framings that I wouldn’t have arrived at on my own, and that end up in my final description.
Keywords and Categories
Amazon allows you to choose seven keyword phrases and two categories for each book. These choices significantly affect discoverability.
AI is useful here for generating candidate keyword lists — phrasing that potential readers might actually search for. I generate a long list with AI assistance and then refine it by checking actual search volume on Amazon (you can do this by typing phrases into the Amazon search bar and seeing what autocomplete suggests — a rough but useful proxy for search frequency).
The Income Reality: Honest Numbers
I want to be honest about this because the internet is full of overstated claims about eBook income and I think honesty serves people better.
My four books generate income. Not life-changing income — not the “passive income” fantasies that some people sell. But real, consistent supplemental income that adds meaningfully to my retirement finances.
Here’s what I’ve learned about realistic eBook income on Amazon.
Volume and pricing: Short practical guides (50 to 80 pages) priced at $2.99 to $4.99 generate a $2 to $3.50 royalty per sale. Longer books (100 to 120 pages) priced at $7.99 to $9.99 generate $5.60 to $7 per sale.
The first three months are the hardest. Without reviews, Amazon’s algorithm doesn’t surface new books to readers. Building initial reviews — through your existing audience, your email list, or a small beta reader group — is the most important early-stage activity.
Your existing audience matters enormously. My first book sold better than an unknown author’s first book would have because I had readers from this site who were already interested in the topics I write about. Starting a platform — even a small one — before you publish significantly improves the income trajectory of the book.
Multiple books compound. A reader who buys one of your books and finds it useful will look for your other books. My third and fourth books generated sales of my first two that wouldn’t have existed otherwise. The income curve for multiple books is meaningfully steeper than a single title.
Realistic expectation for a first eBook with limited existing audience: $200 to $800 in the first year. Realistic expectation for a fourth eBook from an author with an established audience in their niche: meaningfully higher, compounding with the back catalog.
This is not a get-rich strategy. It’s a meaningful supplemental income stream that, importantly, keeps generating income from work you’ve already done — which has a different quality than trading time for money.
What I’d Do Differently Starting Over
Four books in, there are things I’d do differently if I were beginning again. Since this article is meant to be genuinely useful, I’ll share them.
I’d spend more time on step one. The validation and clarity work before the first outline. My first book took longer than it needed to because I started writing before I was fully clear on who specifically I was writing for and what specific gap I was filling.
I’d build a beta reader group earlier. Having 20 to 30 people from your target audience read a near-final draft and give you specific feedback is invaluable. It catches problems that you’ve become too close to the content to see, and it generates your first reviews from people who genuinely read and valued the book.
I’d invest more in the cover earlier. I spent too little on the first cover and too much time later improving it. The cover investment is front-loaded — get it right the first time.
I’d price higher from the start. I initially priced my first book at $2.99 because I was uncertain about its value. The $2.99 price signaled that uncertainty to potential readers. Moving to $7.99 for subsequent books produced better sales, not worse — because the pricing communicated confidence in the value.
Summary and Key Takeaways
Creating eBooks with AI is not about having AI write your book. It’s about a specific division of labor: you provide the knowledge, experience, and perspective that only you have; AI structures, expands, and accelerates the production of everything around that core.
The process, done correctly, produces genuine books — not AI-generated content that a reader will feel cheated by. The books I’ve published contain real knowledge from real experience. AI made it possible to produce them in weeks rather than months and to maintain the pace of publishing that builds a back catalog.
The income is real and compounding. The timeline is honest — months to meaningful results, not weeks. And the process is reproducible by anyone who has genuine knowledge worth sharing and is willing to do the human work that no AI tool can replace.
10 Key Tips for Creating eBooks with AI
1. Start with human clarity, not AI. Before any AI tool touches your book, answer the four validation questions: What do I know that most people don’t? Who specifically is this for? What do they most need? What will they be able to do after reading it?
2. Write your own rough outline first. Then use AI to refine and expand it. The structure that comes from your understanding of the subject is more valuable than the one AI generates without your input.
3. Write your core knowledge in notes before each section. Your personal experience and specific insights are what AI cannot provide. Capture them first, then let AI develop around them.
4. Always rewrite AI-generated content in your voice. Read every paragraph aloud. Anything that sounds like AI, sounds generic, or sounds like it could have been written by anyone — rewrite it in language that sounds like you.
5. Use AI for specific editing tasks, not general editorial judgment. Ask: where will readers lose engagement? Where is this repetitive? What’s unclear? The answers give you a specific editing agenda.
6. Build a beta reader group from your target audience. Their feedback before publication catches problems you can’t see and generates genuine early reviews.
7. Invest in your cover. Test at least two options with people from your target demographic before committing. The cover does more sales work than any other single element.
8. Write your Amazon description before you finish the book. The discipline of writing what the book promises, before you’ve written it, clarifies what you’re actually trying to accomplish and often improves the book itself.
9. Price to reflect confidence in the value. Underpricing signals uncertainty. For a book that solves a real, specific problem for a specific audience, price in the $7.99 to $9.99 range unless there’s a specific strategic reason not to.
10. Start your second book before your first one is fully launched. The income from eBooks compounds across multiple titles. A single book is a project. A catalog is an income stream.
Han Jong-woo is a 71-year-old author and founder of SummitSelect.org. His four eBooks — AI Made Simple for Seniors, AI for Meaningful Living After 60, The $1,000 AI Side Hustle, and AI and Your Health After 60 — are available on Amazon.
Tags: Creating eBooks with AI | How to Write an eBook | AI eBook Publishing | eBook Income | Amazon KDP Guide | Writing eBooks for Beginners | AI Writing Tools 2026
© 2026 SummitSelect.org — All Rights Reserved
It’s never too late to learn, grow, and discover new possibilities.
Explore the books of Han Jong-Woo, where artificial intelligence, lifelong learning, personal growth, andIt’s never too late to learn, grow, and discover new possibilities.
Explore the books of Han Jong-Woo, where artificial intelligence, lifelong learning, personal growth, and meaningful living come together to inspire your next chapter.
Whether you’re curious about AI, seeking purpose, or simply looking for fresh inspiration, you’ll find books designed to encourage, empower, and enrich your life.
▶ Visit Han Jong-Woo’s Amazon Author Page and discover all of his books today. meaningful living come together to inspire your next chapter.
Leave a Reply