I Duplicated My Brain.

Think about the last argument you had with someone close to you.

You remember exactly what happened. What they said. What you said. Who started it. Who was being unreasonable.

Now ask them to tell you their version.

It won't match yours.

And you might think one of you is lying. But no. Neither of you is actually remembering. You're both reconstructing. Pulling fragments and filling the gaps with a story that protects the version of yourself you believe in today.

This is not a flaw in and of itself. This is how memory works biologically. Your brain stores a narrative about facts, edited slightly every single time you retrieve it, filtered through who you currently believe yourself to be.

Before you think I’ve somehow channeled my inner Esther Perel and turned this newsletter into couples therapy.. relax. We’re not talking about relationships today. I’m not qualified, and more importantly, I don’t need my exes emailing me.

The argument was just the entry point. What's actually happening underneath it is something more interesting, and far more consequential.

Here’s what I mean: at the root of every argument, every decision, every belief you hold about yourself, there is data. A finite amount of it. Some was stored. Most was discarded because it didn't fit the story your brain was already telling.

You are navigating your entire life on a unreliable, self-serving reconstruction of reality.

So is everyone else.

-

Now here's where it gets interesting.

I’m certainly not the only one who figured this out.

And those who did, you invited them into your home years ago.

Google. Meta. Your bank. Netflix. Amazon. ChatGPT.

Look at their business models carefully. Most of their core products are free. Why? Because the product was never the product. You are the product. More specifically, your data is.

These companies understood something most people still haven’t absorbed yet: data is the most valuable thing that exists. Not oil. Not gold. Data. Because data predicts behavior. And whoever can predict your behavior can influence it, monetize it, and sell it back to you in a thousand different forms.

They have structured records of everything you do. Every search. Every scroll. Every purchase. Every pattern. They know things about your behavior that you have never consciously noticed about yourself.

And you? You have memory. Which we just established is a story your brain edits to protect your ego.

The conversation about this usually goes: "We should stop giving data to these companies."

If that’s what you just muttered, let me stop you right now. Wrong problem.

The problem isn't that they have your data. They have it, they'll keep it, and opting out is largely theater.

The real question is why they know more about your patterns than you do.

Your own thinking patterns. Your own blind spots and behavioral loops and the gap between what you think drives your decisions and what actually does.

Imagine owning land that sits on top of one of the largest oil deposits ever discovered. Strangers show up every day, drill, extract, and leave with tankers full, and pay you with free snacks. You never thought to drill yourself.

That’s what most people are doing with their own data.

Whoever understands their own patterns first gains something no one can sell you and no one can take: the ability to predict yourself.

-

So I did the opposite of what everyone tells you to do.

I went all in. Balls to the wall.

You probably know what I mean.

I fed it everything. My fears. My 2am spirals. My business decisions. My relationship patterns. The lies I told myself when I didn’t want to look too closely. The versions of me I would never say out loud to another person.

Every uncomfortable question I was too proud to ask a therapist. Every meltdown I disguised as a strategic pivot. Every time I thought I wanted one thing and suspected I wanted another.

I used it the way a confessional works, except this one remembered everything, had no judgment, and was available at 3am.

Fun fact: I wrote to you about it a year and a half ago.

My argument was to use AI as a thought partner. Not to write your emails, not to summarize documents, but to excavate things you couldn't reach alone. To ask it questions that made you uncomfortable.

Most of you thought it was interesting. Few took it seriously.

I kept going anyway.

39 months later brings us to last week:

3,848 conversations. An estimated 70 million words. The entire Lord of the Rings, written 140 times over; except every word was about me, my decisions, my patterns, my blind spots.

I exported all of it.

Every conversation from December 2022 to today. Every message. Every thread. Every voice note. The business decisions and the 2am spirals and the newsletter drafts and the moments where I was trying to figure out who I actually was underneath everything I'd built.

Then I handed it to another AI and gave it a specific prompt.

I asked it to become the world's best expert on me, by doing an exhaustive multi-layered analysis of the entire archive:

• Chronological mapping of my life by phase.

• Core identity and personality profile based purely on evidence.

• Intellectual and professional evolution.

• Behavioral and workflow patterns.

• Limiting beliefs, and how they changed over time.

• What I think I want versus what the data shows I actually pursue.

• How my thinking style evolved. What I adopted, what I abandoned, what stuck.

I asked it to be brutally honest. To cite specific conversations and dates when making strong claims. To look for contradictions. To not flatter me.

What came back was a report. 14 pages, 6,000 words. A mirror. A correction of things I had believed were true about myself, some for years.

While I won’t share the full document with you, here’s the overall structure so you can understand how powerful this can be for you:

Section 1 — Chronological Mapping Five phases. Each one a different version of me trying to solve the same problem with different tools and a different name.

Section 2 — Core Identity & Personality Profile Thinking style. Core values. Recurring emotional patterns. How I make decisions under uncertainty. What actually excites me versus what I think should.

Section 3 — Work & Intellectual Evolution Every field I went deep on. The recurring project patterns. How my standards changed. Key influences and what they actually gave me.

Section 4 — Behavioral & Workflow Patterns Energy cycles. How I learn. Collaboration style. Goal-setting tendencies. Information consumption.

Section 5 — The Living Profile Top 5 superpowers. Top 5 blind spots. 10 memory anchors — quotes pulled from 3 years of conversations that, put together, are the most honest portrait of me I've ever seen.

It ends with recommended future directions, which I am already applying.

What the data actually told me, and why it matters for you.

A few things stood out because they were precise in a way I couldn't have been on my own.

1. What I said I wanted and what I actually pursued were two different things.

This one I didn't see coming.

The report went through three years of decisions and mapped the gap between my stated priorities and my actual behavior. Not what I claimed to value. What I repeatedly chose when it cost me something.

The delta was a little uncomfortable. The version of myself I described to people and the version that showed up in the data weren't the same person.

And I guess a lot of people (like it was the case for me) never see this gap clearly. They operate on the story. The data doesn't care about the story.

-

What do you actually pursue when no one's watching and nothing's framed as a test?

2. I make decisions completely backwards.

Big, life-changing decisions (leaving a business with 37 employees, moving countries, writing a book) I made fast. Days, sometimes. The data showed these were almost always right.

Small tactical decisions (an opening line of a newsletter, a pricing tier, a subject line) I agonized over. Rewrote the same introduction six times when it was already good on the first draft.

I knew this vaguely. I had never seen it written down, dated, evidenced across 39 months. A vague feeling about yourself is very different from a verified pattern with timestamps.

-

Where are you spending the most decision-making energy? Is that the right place?

3. How I thought I learned best was wrong. And it was costing me.

The report went through three years of intellectual breakthroughs and mapped where they actually came from.

Not the courses I bought. I bought several. Never finished one.

Not the books I read cover to cover. The ones that moved me were ones I consumed in pieces, immediately applied, and stopped when they stopped being useful.

Every real shift came from the same sequence: experience something, try to write about it, understand it through the act of writing.

The newsletter isn't my product. The report confirmed it's my thinking process made public.

-

What's the last thing you learned that actually changed your behavior?

A dataset is like a knife. You can cut it reality many ways.

When you have 39 months of raw data, the report is just one way to read it. So I ran a second analysis. Not on what I said, but on how I said it. The specific words I used. How they changed year over year. What I stopped reaching for. What replaced it.

Turns out your vocabulary is a cleaner mirror than your memory. You can't fake word frequency across 70 million words.

A few things the language showed that I couldn't have seen otherwise:

The moment my real estate business closed, the word "French" spiked to its highest point in the entire archive. I hadn't noticed I was doing it, but when the professional identity collapsed, I unconsciously anchored to the one thing that couldn't be taken: where I came from. It lasted about six months. Then it faded too.

The word "authentic" exploded in 2024, peaked, then nearly vanished by 2025. Not because I abandoned it, because I absorbed it. Words you stop needing are often the ones that worked.

And here's what the full arc looks like visually.

1. Struggle peaked in 2024

The deepest year of the work. By 2026, mastery overtook it for the first time:

2. Topic Lifecycle

"Identity" goes from 2 occurrences in 2023 to 1,273 in 2025. You don't choose what obsesses you. But you can choose what to do with it.

This is what 39 months of thinking out loud leaves behind. Not a journal. A dataset. And the dataset doesn't lie:

Why I am sharing this with you.

The conversation you are already having (wherever you think out loud, write, process, talk to an AI) is a potential dataset. You are already generating the raw material. The question is whether you're treating it as disposable or as something to return to.

The brain stores stories. Data stores patterns. You need both. But right now, almost everyone is running entirely on the story which is edited, biased, and optimized to protect your current self-image rather than to show you what's actually true.

You don't need 3,848 conversations to start. You need a practice of thinking out loud in a place you can return to. Consistently.

The aggregate is what becomes valuable. One conversation is a thought. Three years of conversations is a map of your mind.

Most powerful companies in the world built empires on your data.

I built a second brain on mine.

The asymmetry doesn't have to run in one direction.

If you want the actual prompt I used to generate this report, it’s available here.

Stay smart,

Benoit

P.S. — The Double Entry Protocol is the system I built when I realized understanding my patterns wasn't enough to change them. The people running it right now: entrepreneurs doing $10M+, coaches, therapists. If that's your world: You can close the gap here.

P.P.S. — None of this would exist if I hadn't gotten out of my own head first. That's what Unf*ck Your Thinking is about. [Amazon] [PDF]