The method
A machine writes.
A doctor checks.
“James Whitfield” is an artificial intelligence — a large language model. He writes the books. We tell you that on the first page, on purpose, because the rest of the shelf won't.
Why say it out loud?
The wellness shelf is already full of books written by machines — sold under invented author names and tearful origin stories, with no one in the process paid to care whether a word of it was true. The dishonest thing was never using the machine. It was the hiding, and the not-checking.
So we built it the other way around: loudly machine-written, obsessively human-checked. James writes. A practising doctor verifies. Neither of us ships a sentence the other can't answer for.
The moment that proves the process
Early on, James handed over a citation — author, journal, year, a clean-looking reference behind a claim about memory and the vagus nerve. The kind you'd skim past. But the rule is: open every reference and read enough to confirm it says what we say it says. So the doctor went looking for that paper.
It didn't exist. Not paywalled, not hard to find — never written. If no human had been standing there, that invented study would have gone into the book — and you'd have had no way to know. That's the whole difference between this book and the one you almost bought.
How every James Whitfield book is made
Most AI books are made end to end: someone types a prompt, walks away, and publishes whatever the machine hands back. That's how the shelf filled up with slop.
We build the opposite way — middle to middle. A human stays at both ends. The human owns the start: the vision, the structure, the arguments, the judgement about what's true and what matters. The machine does the middle: the heavy lifting of turning a detailed plan into prose. And the human owns the end: editing every line, checking every claim against its source, and signing off. The machine never gets the first word or the last.
That isn't a compromise — it's the quality. The small slice that stays human, the framing at the front and the judgement at the back, is the slice that decides whether a book is worth your time. Everything published under this name is built this way, or it isn't published.
The standard, in the open
- Real primary sources. Every factual claim is traced to a study, a clinical guideline, or a textbook — never to another popular book or a machine's confident summary.
- Every citation is opened and read. Author, journal, year — and the only question that matters: does the paper actually say what we claim? That check caught a study the AI had invented.
- Honest numbers, not impressive ones. Where a device's headline figure was 71% and its rigorous trial figure was 35%, we printed 35%.
- Fringe voices stay out. However viral, claims from outside the genuine evidence base don't appear — even when they'd make a better story.
- We hired our own enemies. Before you saw it, a panel built to destroy the manuscript — a sceptical scientist, a specialist we criticise, a hostile reviewer — went hunting for what we'd missed.
- A human is accountable. Not a generated face. A real, reachable doctor who fixes what's wrong — and whose judgement you're trusting, not the machine's.
About the human
The reviewing adviser is a practising, registered doctor who fact-checked every claim against its primary source and is accountable for what made it onto the page. By choice, they work under the role rather than a personal byline — because what matters isn't whose name it is, but that a real, qualified person stood between the machine and you, and checked.
Machine-written. Human-true.