No, using AI to write blog posts isn’t cheating, but how you use it decides whether the result is yours. I have ADHD, roughly a billion ideas a day, and an AI setup with strict writing rules I built myself. It interviews me before it drafts a word. Every opinion, story, and fact in my posts comes from me, and I make the final call on every line. The thinking is the writing, and I still do all of it.
The shaming is real
Spend ten minutes in any writing community and you’ll find it: someone mentions AI helped with a post, and the replies pile on like they admitted to copying a stranger’s homework. The words slop, lazy, and “not a real writer” get thrown around fast, usually at people whose actual process nobody asked about.
I don’t feel guilty about how I write, and I refuse to. But I know plenty of bloggers are second-guessing themselves because of those threads. This post is for them, and it’s also me putting my own process on the table so you can judge it for yourself instead of judging a caricature.
Is it cheating to use AI to write blog posts?
It’s not cheating, because the thing readers come for was never the typing. They come for what you know, what you tried, what broke, and what you’d tell a friend to do. If those come from you, the post is yours no matter what tool typed the sentences. Nobody calls a podcaster a fraud for not soldering their own microphone.
What CAN go wrong is skipping the you part entirely. Type a keyword into a chatbot, say “write a blog post about composting,” publish whatever falls out, repeat 50 times. The result is slop, and the internet has plenty. But notice what’s missing from that picture: any input from a human with actual experience. A post with nobody home reads like a post with nobody home, whatever tool typed it.
What my ADHD brain does with a post idea
I was diagnosed with ADHD at 46, which explained about four decades of having no less than a billion ideas a day and no reliable way to catch them. Ideas show up while I’m washing the dishes, mid-shower, in the middle of building something else. By the time I’d sit down to outline one the old way, three new ones had shoved it out the door.
So my writing process starts with what I can only describe as word vomit. I dump everything in my head at the AI: the idea, the half-formed opinions, the tangents, the typos, all of it. Then it helps me make sense of the pile. For a brain like mine, this is the difference between a billion ideas dying in my head and some of them becoming actual posts.
The organizing was always the wall. It was never the ideas, and it was never having something to say.
Same brain, different fix: I wrote about why a body double gets me started on the tasks I avoid. And if a blank page is your wall rather than too many ideas, I reviewed a journal built for exactly that freeze.
So did the AI write it, or did I?
Yes, the AI types most of the sentences. Any softer version of that falls apart the second anyone pushes on it. And I still call the posts mine, because typing sentences is the smallest part of what writing a post is.
What I bring: the idea, the experiences, the opinions, the real numbers, and the final call on every line. Before any draft exists, Claude Code extensively interviews me about the post. One question at a time, digging for what happened, what I think, and what I’d push back on. My answers become the raw material. The AI also carries a stack of writing-rule skills I built and update constantly, down to a list of banned words and phrases long enough to be a little embarrassing.
Then I’m hands-on through every revision, and I’m the only one who hits publish.
This exact post got made this way. It started as a messy paragraph I typed at Claude Code, followed by an interview about where the line sits, how much of my diagnosis to share, and what I’d say to a skeptic. If you want the receipts, the interview is baked into the structure you’re reading.
AI slop vs. AI help: where’s the line?
The line is input. Slop starts from a keyword. Help starts from a person with something to say. Side by side:
| Step | AI slop | AI help (how I do it) |
|---|---|---|
| Starting point | A keyword and a word count | An idea I’ve been chewing on all day |
| The prompt | “Write a blog post about X” | An interview where I answer 6 to 10 questions first |
| The facts | Whatever the model half-remembers | My own results, checked against real sources |
| The opinions | Averaged internet mush | Mine, including the unpopular ones |
| The voice | Default chatbot cheerful | My rules, my banned words, my weird asides |
| The edit | None | Me, on every line, sometimes grumpily |
| Publish | Automatic | Me, manually, every time |
One column produces a post nobody needed. The other produces my post, faster and better organized than my brain would’ve managed alone.
The rules my AI has to follow
Guardrails are what make the help safe to lean on. Mine:
- It never publishes. Every post lands as a draft and I push the button.
- It never invents a story or detail about my life. No sentence from me means it doesn’t go in.
- It never states a fact without a real source. Specs, prices, and numbers get checked against the live thing.
- It interviews me before every single post, not just the big ones.
- It follows my voice skills, which I tweak so often they should charge me rent.
- I reject what doesn’t sound like me. About half the social copy my automations generate never gets posted, because I read all of it.
If you’re wondering what those voice skills look like, I’ve written up how I customize every skill I use and how to build a voice skill of your own.
What to say when someone says “the AI wrote it”
You don’t owe anyone a defense, but if you want one, keep it short and specific: “The ideas, the opinions, and the experience are mine. The AI helped me organize and draft, and I edited every line. Which part do you think a machine came up with?”
Then let it go. You know what your process looks like from the inside. A drive-by commenter doesn’t, and there’s no arguing someone out of a position they scrolled into.
Get interviewed before you draft
The interview step is what I’d hand any blogger who wants AI help without the slop, so I built a little tool for it. Answer four blanks about what you write and it builds you a custom Claude Code skill, a paste-into-any-chat prompt, or both. Yours makes the AI interview you, one question at a time, before it drafts a word.
Nothing you type is stored or sent anywhere, and there’s no signup.
Frequently asked questions
Does Google penalize AI-written blog posts?
No. Google’s guidance says it rewards helpful content regardless of how it’s produced. What its spam policies target is scaled content abuse, meaning piles of low-value pages pumped out to game rankings. A post built from your real experience with AI help isn’t in either bucket.
Do I have to tell readers I use AI?
For a personal blog, no rule requires it. I talk about it openly because building with AI is what this blog covers, and because I think hiding it feeds the exact stigma this post is about. Your call.
Will readers be able to tell I used AI?
If you hand it a keyword and walk away, yes. Default AI writing has tells, and readers are getting better at spotting them. If your ideas, stories, and edits are in the post, what they read is you.
How do I use AI without losing my voice?
Give it your rules instead of hoping. A written style guide, a banned-words list, and real examples of your writing get you most of the way. Making the AI interview you first keeps the raw material yours.
Is it lazy to use AI for blog posts?
No. The work moves, it doesn’t disappear. I spend my writing time answering interview questions, checking facts, and editing, instead of staring at a blinking cursor trying to organize a billion thoughts by hand. For my ADHD brain, the old way wasn’t more noble. It was just slower and produced fewer finished posts.
A tool I built
Subscription Graveyard
Bury the subscriptions you’ve cancelled, give each one a tombstone, and watch what you’re saving pile up.
Bury your first one →I run an affiliate program on the tools I build. Approved affiliates earn 20% on what their referrals pay, for up to a year. US only for now. See the program →
Related reading
- I Don’t Install Claude Code Skills, I Customize Them
- Stop Pasting Your Style Guide Into Every New Chat
- Turn Your Blog Post Checklist Into a Claude Code Skill
- How I Customized My LLM to Write in My Voice
AI is here to stay, and there’s no putting the toothpaste back in the tube. You can sit out the tool and keep hand-organizing every draft, or you can make it work your way, with your rules, on your ideas. I picked my way and I’m not looking back.
Subscribe if you haven’t already, and go make an AI ask YOU the questions for once.
