You think this is AI. Fair enough, I’ve used AI to help write the page explaining why it isn’t the one calling the shots here. Stick with me, that’s the whole joke, and everything below just proves it.
THE BORING, HONEST LIST
Right, let’s not be precious about this. AI drafts. It researches faster than I can, digs through studies while I’m still finding my glasses, and gets a first version of a blog post or a script onto the page instead of me staring at a blank one. It helps with the image prompts too, the watercolour illustrations across the library, drafted with AI, then checked against what I actually meant.
It edits. Catches the sentence I’ve written four times because I keep circling the same point. Proper second pair of eyes, the sort you’d want on a Tuesday afternoon when you’ve been staring at the same paragraph for forty minutes and it still isn’t landing.
What it doesn’t do is decide what’s true, what goes out, or what I say about my own body. That bit’s still entirely mine.
WHAT IT CAN’T DO
NHS blood tests going back to 2012, since 2012. Coming off four medications, my decision, in February 2024, GP told after the fact, not before. No AI wrote that. No AI lived it either. That’s not a prompt, that’s a decade.
Unbroken NHS Blood Record,
Every One Logged
Since 2012.
Medications Stopped
My decision, February 2024.
Told my GP after, not before.
Person, One Body, Thirteen Years
N=1. Me, my blood tests,
my GP. Not a clinical trial.
Every moment I talk about, the mirror moment, the medication decision, the bloods that made my GP raise an eyebrow, traces back to something real. An actual date. An actual result. Not a vague memory dressed up for a script.
Before any of that gets reused in something new, it gets checked against the original. So when I say “this happened in month three,” it’s because month three is where it happened, not because it sounded about right.
I don’t cite papers to sound clever. I use them for one job, understanding the mechanism behind what I’ve already lived.
NHS and NICE guidance comes first. Proper studies and reviews come next. Researcher commentary, books, podcasts, other people’s interviews, comes last, and it’s always labelled as someone’s view, never dressed up as settled fact.
When I bring in a Nature journal article or a study, it’s never to replace my own results. It’s to explain why they happened.
THE BIT, TRANSLATED
Here’s a real flavour of it, roughly what comes back when a script or an article goes off for a check.
What the AI actually said: “This contradicts the position you took in March, and it’s going to land badly with the GP-cohort readers.”
Translated from AI-diplomatic into actual English: “You’re wrong, and everyone’s going to notice.”
Not rude, exactly. It just doesn’t let things slide, which is more useful than something that only tells me what I want to hear.
Nothing gets published, recorded, or sent until I’ve read it myself.
WHAT YOU DON’T SEE
What ends up on the site is the tidy version. Underneath it there’s a fair bit more, drafts that got binned, arguments over whether a sentence was too soft or too sharp, stuff that got cut because it wasn’t quite honest enough, or wasn’t quite Neil enough.
Most of that never sees daylight. Some of it does, in the newsletter, where I’m a bit less careful about tidying things up.
Join the email list. It’s where the rougher, less edited version of all this actually lives.
This is educational content, not medical advice. If you have a health condition or take medication, speak to your GP before changing your diet, fasting pattern, or lifestyle.
