How one author used AI to write 370 books and make $1 million in 20 months
Coral Hart built an AI system to publish 370 books in 20 months. She made $1 million this year, proving AI is a tool, not the creator.
Business · Source: Dubai Telegraph (AFP wire)
What happened
South African author Coral Hart published 371 books in 20 months using AI. She operates under a pseudonym to protect her identity. The internet called her a grifter and sent her death threats. She ignored them and kept building. This year alone, she sold over one million books across digital, print, and audio formats, clearing $1 million in revenue.
Hart spent 14 years in traditional publishing before pivoting to AI in February 2025. She spent three months building a custom framework. She does not just prompt a model to write a book. She defines the concept, style, character arcs, and chapter outlines. Then she routes specific tasks to different models.
ChatGPT handles dialogue. Claude manages story structure. Grok writes the spicy scenes. The first draft is usually rough. Hart runs it back through AI for developmental edits to fix plot holes and continuity errors. Finally, a human proofreader scrubs the text for AI tells. Now, she is turning her internal tool into a commercial software product called Authorythm.
Key facts
- 371 — Number of titles Hart published in 20 months
- $1 million — Hart's estimated earnings so far this year
- 14 years — Time Hart spent in traditional publishing before using AI
- 3 months — Time it took Hart to build her AI writing framework
- Authorythm — The name of the software she is launching for AI authors
Why it matters
This shatters the myth that AI will replace human taste. Hart proves that AI is just a high-speed pen. If you ask a model to write a book from scratch, you get trash. The real moat is human curation, market research, and structural design. Builders need to stop trying to automate the entire creative process. Focus on automating the execution while keeping the human firmly in control of the vision.
The second-order effect is the commoditization of niche content. Hart targets specific reader demands with rigorous market research. As more creators adopt frameworks like Authorythm, the volume of hyper-targeted media will explode. The winners will not be the models. The winners will be the operators who can orchestrate multiple specialized AI agents to ship products faster than traditional gatekeepers can blink.
For builders
Route tasks to specialized models
Do not rely on one model for everything. Hart uses ChatGPT for dialogue, Claude for structure, and Grok for edge cases. Build routing layers in your apps to leverage the specific strengths of different LLMs.
Productize your internal workflows
Hart trained hundreds of people on her method and is now launching Authorythm. If you build a successful internal tool to solve your own bottleneck, package it and sell it to others in your industry.
Human-in-the-loop is mandatory
AI cannot finish the last mile. Hart uses human proofreaders to fix AI tells and rewrite clunky syntax. Budget for human QA in your AI products or your users will reject the output.
My take
Everyone is crying about AI ruining art, but readers do not care. They buy what entertains them, and they return garbage. I love seeing a founder ignore the noise, build a system, and ship product until the market pays them a million dollars.
Original reporting: Dubai Telegraph (AFP wire). This is my rewrite and opinion.