Clone your brain into Claude

Most people treat AI like a vending machine

June 08, 2026

In partnership with

I keep running into smart people who type a quick request into Claude, get a flat generic answer, and walk away convinced the whole thing is overrated. I used to nod along. Then I realized the problem was never the model.

The model has no idea who they are, how they think, or what they actually want. So of course it sounds like everyone else. It's guessing.

That's why a step-by-step breakdown I read on LinkedIn stopped me cold. The creator laid out a full system for copying your entire brain into Claude, so the model stops guessing and starts sounding like the real you. I was impressed enough to run it myself and break the whole thing down here.

The big idea is simple. You build three personal files one time, then every AI you touch reads them and instantly knows your voice, your goals, and your taste.

2026 State of AEO Report

A year ago, most marketers weren't thinking about AI search. Now it's one of the fastest moving channels in the industry and nobody has a playbook yet.

So we built one. We surveyed hundreds of marketers to find out how they're approaching answer engine optimization, where they're investing, what's actually working, and what isn't.

The result is the 2026 State of AEO Report. Real data. Real strategies. A clear picture of where AI search is headed and how to get ahead of it.

*Ad

The 11-step build

Each step has a clear job. I added the reasoning so you know why it matters, not just what to click.

  1. Set up Claude properly. Download the desktop app, set the top Opus model as your default, and turn on Extended Thinking so it reasons before it answers.

  2. Make a home for your brain. Open Cowork mode (the tab between chat and code) and create a folder called "Brain" inside. This is where your identity will live.

  3. Install voice-to-text. Grab a free tool like Wispr Flow. The author argues speaking is faster and more honest than typing, because typing makes you edit yourself.

  4. Run a 100-question taste interview. Have the AI ask you 100 questions across 7 categories. Push past every vague answer and get specific.

  5. Compress the dump. Turn that giant brain dump into a clean summary and save it as about-me.md.

  6. Write your bans. Create an anti-ai-writing-style.md file listing every word and habit you refuse to see, like banning "delve" and killing the "it's not X, it's Y" pattern.

  7. Write your north star. Build a my-company.md file with your goals, your focus, and your hard no's. The creator calls this the north star file.

  8. Test it in a blank chat. Open a fresh chat and write a prompt only you would write. If the output sounds like you, you're on the right track.

  9. Load the folder. Drop all three files into your "Brain" folder so Claude reads them on every single turn.

  10. Port it everywhere. Copy the same files into ChatGPT, Grok, and Gemini, so you get the same brain in every AI.

  11. Edit it forever. Install Obsidian (free), open your folder as a vault, and update the files as you grow.

Why voice beats typing

This was my favorite part of the post. When you type, you polish and hedge. When you talk, the truth slips out.

Dictating your context captures how you actually think, including the messy, opinionated parts that make your voice yours. So instead of writing careful little answers to those 100 questions, you just talk and let the tool transcribe. I tried it both ways, and the spoken version was unmistakably more me.

PRDs by voice. Bug reports by voice. Ship faster.

Dictate acceptance criteria and reproductions inside Cursor or Warp. Wispr Flow auto-tags file names, preserves syntax, and gives you paste-ready text in seconds. 4x faster than typing.

*Ad

The anti-style file is the secret weapon

Most people skip this one, and the creator says that's exactly why their AI output reads like every other AI on the planet.

Your anti-ai-writing-style.md file is a list of bans. Words you hate. Phrases that aren't you. The fake-deep contrast patterns. As the post put it bluntly: without this file, Claude writes like Claude. With it, Claude writes like you.

A few rules you can borrow to start your own:

  1. Ban filler words you'd never say out loud.

  2. Kill negative parallelism, the "it's not just X, it's Y" formula that screams robot.

  3. Note your real sentence rhythm, short and punchy or long and winding.

Your north star file

The third file, my-company.md, keeps the AI pointed at what matters right now. The author suggests keeping it over 1,000 tokens and updating it once a quarter.

It holds your current goals, your main focus, and the things you're actively saying no to. When the AI knows your priorities, its suggestions stop being generic and start being useful. This is the file I underestimated at first, and it ended up doing the most work.

Build once, use everywhere

Here's the payoff that makes the whole effort worth it. Once those three files exist, they're portable.

You can upload the same set to ChatGPT, Grok, or Gemini and get a consistent version of your brain across all of them. You can even hand the files to a teammate or a ghostwriter so they can write in your voice too. You build it one time and reuse it forever.

A few uses worth trying:

  1. Draft emails and posts that already sound like you, with far less editing.

  2. Onboard a new AI tool in seconds by pasting in your three files.

  3. Give your team a shared "voice of the founder" so messaging stays consistent.

  4. Keep every AI aligned with your current quarterly goals.

The line I keep thinking about

The one sentence from the post that stuck with me: the mystery you've built your identity on is usually just being vague.

That landed because it's true for me. Writing these files forces clarity, and clarity is what makes the AI finally get you. The exercise isn't really about the AI. It's about pinning down what you actually think, in writing, for the first time.

If you've ever felt like your AI output sounds flat and faceless, this is the fix. Block out an hour this week, open a blank chat, and start with the taste interview. By the end you'll have an about-me.md worth keeping, and an AI that finally sounds like you.

“Who is this person again?”

You’ve had that moment. Walking into a call, scrolling through old emails, trying to remember what you promised. Lindy texts you a brief 15 minutes before: attendee context, past discussions, open items, talking points. All pulled automatically. Try Lindy free.

*Ad