A book-length report, researched and written by AI agents: how Robert Greene makes his books, what it would take to rebuild that process with AI, and what a bad actor could do with the same machine.
Chapter 1··22 min read
What this report asks about Robert Greene and AI, how it was researched from 22 interview transcripts and a stack of written sources, what that evidence can and cannot show, and how to read the fifteen chapters.
Chapter 2··45 min read
Robert Greene published nothing until he was 39. This chapter follows the forty years before The 48 Laws of Power: Los Angeles, Greek at Berkeley, journalism, Europe, dozens of jobs, Hollywood, the theatre, and a walk in Venice with Joost Elffers.
Chapter 3··41 min read
How Robert Greene finds, sorts and reads the hundreds of books behind each of his own: the numbers he gives, the shelves he hunts on, the footnotes he follows, the pen he reads with, and how it compares with the way researchers are trained to read.
Chapter 4··45 min read
Everything on the record about Robert Greene's handwritten notecards: the wait before carding, how many cards a book gives, the colours, the boxes, how chapters come out of them, the versions Ryan Holiday and Billy Oppenheimer run, where the idea comes from, and what an AI version would have to keep.
Chapter 5··44 min read
How Greene turns a box of notecards into chapters, how he hunts for the story that opens each one, and the story, moral and interpretation template he has used since The 48 Laws of Power.
Chapter 6··41 min read
What happens after Greene's first draft: twenty drafts, the cutting, his rules for voice, his two readers, the deadlines that nearly sank two books, and the daily routine that held it together before the stroke.
Chapter 7··51 min read
Each of Robert Greene's eight books, one at a time: why he wrote it, what he read, how long it took, how it's built, how it sold and what the critics said.
Chapter 8··35 min read
How a stroke in August 2018 took away Robert Greene's typing, walking and travel, and how he wrote The Law of the Sublime anyway: two notebooks, sticky notes, dictation, printouts and one hand, over seven years.
Chapter 9··40 min read
What Robert Greene has said about shortcuts and AI from 2013 to 2026, how five other writers built their research systems, and which parts of each system a model could take over without breaking what the writer says matters.
Chapter 10··37 min read
A cost model for a Greene-scale book made with AI agents: tokens and model prices for every stage, buying and scanning the books, the human hours that remain, the legal benchmark, and what happens at 10, 100 and 1,000 books a year.
Chapter 11··66 min read
A full engineering design for an agent pipeline that follows Greene's method: ten stages, the data model behind the cards, the critics loop, the human gates, and one chapter walked through end to end.
Chapter 12··47 min read
A survey of the open-source projects, APIs and benchmarks for each stage of a Greene-style pipeline, with verified GitHub stats, the evidence behind each one, and the gaps nothing fills yet.
Chapter 13··46 min read
What the studies say about how often AI research tools invent or misuse sources, what went wrong in the public incidents, and how to design a fact gate, a human audit and an error budget for a 150,000-word book.
Chapter 14··43 min read
A threat model for authors, publishers, platforms and readers. What a fictional bad actor could build by feeding the book pipeline with pirated libraries and removing its checks, what breaks, who gets hurt, what the law says, and how to defend against it.
Chapter 15··10 min read
The three answers in one place: how Greene works, what an AI version can and can't take over, and why the expensive parts are the ones that protect readers.