Turning a book or blog archive into a course with AI

If you have written a book or kept a blog for years, you may be sitting on most of a course already. Hundreds of posts, a few long guides, a book full of hard-won advice: the knowledge is there. The trouble is that it is organized for reading, not learning. Posts were written in whatever order ideas occurred to you, chapters assume someone reading straight through, and nothing asks the reader to practice.

Restructuring all that by hand is a daunting job, which is why so many archives never become courses. AI can take on the sorting, mapping and first-pass restructuring. You bring the thing the archive cannot supply on its own: a clear idea of what learners should be able to do at the end, and the judgment about what to keep.

Check your rights first

Before you reuse anything, make sure you are free to. If you self-published, you probably are. If a publisher holds rights to your book, your contract may limit how you reuse its content in a paid course. Guest posts, co-written pieces and material you quoted from others need a similar check. Copyright and contracts vary, so check what applies to you with a qualified professional.

Inventory what you have

Start by getting the archive into a form an AI tool can read. Export your blog posts or copy the text of each chapter into documents, one file per piece if possible, with the title and publication order. Tools that handle many documents at once, such as NotebookLM, Claude Projects or a custom GPT, are helpful here because an archive can be large.

Tobias has written a fermentation blog for a long time, with several hundred posts on sauerkraut, kimchi, kombucha, hot sauce and more. He started by asking an assistant to list every post with a one-line summary and a suggested theme, then to group the themes. Within an afternoon he had a map of what he had written, including clusters he had forgotten about and topics he had covered five times in slightly different ways.

Find the learning path hidden in the archive

An inventory tells you what you have. A course needs a path. Decide first what a learner will be able to do at the end, then ask the AI to map your material onto that path. If you need help with the destination, writing learning objectives with AI walks through it.

I want to turn my existing content into an online course. The learners are [describe your learners and their starting point]. By the end, they should be able to [describe the finish line]. Below are summaries of my posts or chapters, with titles. [paste inventory]

Propose a sequence of four to six modules that leads to that finish line. For each module, write one learning objective and list which of my pieces could supply the content, in teaching order. Then list: pieces that repeat each other, pieces that do not fit this course, and gaps where the path needs something I have not written. Use only my content; do not invent new topics to fill gaps, just name them.

For Tobias, the path became: safe basics and equipment, vegetable ferments, drinks, sauces and troubleshooting. His archive covered drinks thoroughly and troubleshooting barely at all, even though it was the most common question in his comments. That gap became the most valuable module in the course. For more on this kind of analysis, see finding gaps in your content library with AI.

Rewrite pieces as lessons, not posts

A blog post and a lesson do different jobs. A post informs or entertains. A lesson moves a learner from not being able to do something to being able to do it. When you convert a piece, it usually needs an objective at the top, fewer tangents, a practice activity and a short recap.

Ask the AI to restructure one piece at a time while keeping your words. Give it the objective and the original text, and ask it to reorganize the material into a lesson with a clear opening, the core teaching, one practice activity and a recap, cutting tangents but keeping your phrasing and stories wherever possible. Ask it to list what it cut so you can put back anything important.

Adaeze wrote a book on negotiation for freelancers. Her chapters were full of anecdotes, which made the book enjoyable but the lessons long. The AI suggested moving most of the stories into optional “from the field” sidebars and keeping one per lesson in the main flow. She kept the idea but chose the stories herself, because she knew which ones readers had always mentioned.

Fill the gaps yourself

This is where AI must step back. When the map shows a gap, AI can help you outline a new lesson, suggest a structure or ask you interview-style questions to draw out what you know. It should not write the expertise for you. An archive-based course sells because it is built on your experience; invented content filling the gaps would undermine exactly that.

One productive approach is to ask the AI to interview you. Have it ask ten questions about the gap topic, answer them in a voice memo or typed notes, and then use your answers as the raw material for the new lesson.

Review the whole course, not just each lesson

Once the lessons exist, read the course from start to finish. Look for advice that has changed since you first wrote it, older posts that contradict newer ones, references to things that no longer exist and jumps in difficulty between lessons. Your archive was written over time, and your views probably evolved. The course should reflect what you believe now. For the non-AI basics of sequencing, see structuring a course into modules, and compare your path with an AI-assisted course outline if you want a second view.

First steps

1. Confirm you have the rights to reuse your book or posts in a paid course.
2. Export the archive and ask AI for an inventory with summaries and themes.
3. Write the course finish line, then run the mapping prompt.
4. Review the proposed path, the repeats and the gaps.
5. Convert one piece into a lesson, keeping your words, and check what was cut.
6. Fill gaps from your own knowledge, then read the whole course for contradictions and outdated advice.