For those of you who arrived recently: welcome. This is what Sundays usually look like around here. If you’ve been here a while, I left you a note at the bottom.

I used to optimize everything. In fact, I still hold two world titles for that.

Things like the precise moment I should start unloading my cart at the grocery store.

There was, according to me, an optimal moment: You had to account for how much the person in front of you had left, how quickly the cashier was scanning, how many things were still in your cart, and begin unloading at exactly the right time (and in the right order) so that your last item landed just as the previous customer finished paying. Too early and the groceries sat there waiting. Too late and the cashier waited for you. It also had to be categorized by heavier items first, to start optimizing for the way the items would be placed in the bag on the other side. When someone tried to help me unload the cart, I would get stressed out, because it messed with my optimization.

Needless to say, I was very good at it.

I did the same thing in the car. I couldn't drive five minutes somewhere without trying to find the fastest possible route. I would check different apps, remember which lights took too long, try a side street, compare the result against whatever Google said when I left. Every once in a while I would beat it by 32 seconds.

Which was fantastic, because I now had an additional 32 seconds to do absolutely nothing with.

Usually I would use them to think about whatever I was going to optimize next.

You might have some slightly less compulsive version of that today.

There is something genuinely satisfying about optimization. It gives ordinary life a scoreboard. You find the faster route. You improve the system. You shave three minutes off something. Reality behaves slightly more like you wanted it to, and for a moment you feel like you've won.

It took me a long time to realize that I wasn't really optimizing for time. I was simply optimizing for control.

Around 2021, I wore a Fitbit to bed every night. The first thing I would do in the morning was check how I had slept. Six hours and whatever. Deep sleep bad. Recovery bad. Some number telling me that last night had apparently been a disaster.

And I would immediately think, Fuck. I'm going to feel like shit today.

Then one morning I had a thought that eventually made me stop wearing trackers altogether.

What if I don't?

I had barely been awake long enough to know how I felt. My body hadn't told me anything yet. I had simply looked at something outside of myself, received an interpretation of what had happened inside of myself, and accepted that interpretation before having the experience. Maybe I had slept badly. The Fitbit was probably right about the hours. That's not the interesting part.

The interesting part is that I never got to find out how I would have felt without knowing the number. The interpretation arrived before the experience.

This letter came about when I saw Rich Roll put out a collaboration add with WHOOP.

He checked his WHOOP at the gym, saw that his recovery wasn't where it should be, and decided to take it easy instead. Again, maybe that was the correct decision. Maybe the data prevented him from overtraining. Sometimes people say things because they're being paid to say them in ads, and I can't know which this was.

But I actually think he believed it, which makes the example more interesting, not less.

Because look at the ritual. You look at a device attached to your body, and then decide what your body should do based on what the device said. Somehow we have decided this is knowing yourself better.

At some point the information stopped informing the experience and started determining it. The tool you bought to understand yourself became the thing you consult instead of yourself.

And the strangest part is that from inside the system, this looks like self-mastery.

That's the paradox I keep coming back to with optimization. The more control we acquire, the easier it becomes to give control away to the systems we created to acquire it.

Now, before I get too smug about a grown man asking his wrist for permission to exercise, I should probably mention that I managed to do my own version of this a few months ago. When AI agents got good enough to be useful, I spent 2 full weeks building a bunch of systems around my content to make everything faster and more efficient. Then I improved the systems, built agents to run parts of them, improved the agents, and at some point realized I was optimizing the optimization. I had built a remarkably efficient machine around creating things while spending considerably less time actually creating them. So I got rid of all of it. Turns out that helped.

I think this is also why perfectionism stops so many people from starting. Perfectionism is just another attempt to solve uncertainty before entering it. You want to know that the thing will be good before you make the thing. You want the correct strategy before you start. You want enough information that you can remove the possibility of looking stupid, failing, wasting your time, choosing wrong.

The problem is that most of the information you need is on the other side of doing it badly.

I didn't start writing online because I finally knew how to write online. I started because I became willing to suck at it and publish anyway.

I wasn't suddenly confident. I wasn't certain it would work. I became willing to produce something that might not be good and let reality answer me.

And reality did.

Everything useful I know about writing came after that decision. Which means all the information I had been trying to acquire before acting was sitting behind the action the entire time.

This is where I think optimization, perfectionism and our obsession with control all meet.

We don't like uncertainty. We really don't like not knowing.

So we measure things because measurement makes them feel more predictable. We build systems because systems make outcomes feel more repeatable. We optimize because every variable we bring under control removes one small place where reality can surprise us.

Except it doesn't stop.

Once you optimize sleep, there is recovery. Once you optimize recovery, there is nutrition. Once you optimize nutrition, there are supplements, hormones, training, light exposure, temperature, timing. Once your morning routine is perfect, the afternoon suddenly looks suspiciously unoptimized.

Control doesn't satisfy the desire for control. In fact, it expands the territory available to it.

And optimization is an unusually good machine for producing the perception of control, because there is always a better version of you available to maintain. Healthier. More productive. More disciplined. Better rested. More efficient. More informed. More controlled.

There is no finish line because there is no final version of a person. And if you want to see the logical endpoint of that idea, eventually you arrive at death.

Because underneath all of this is uncertainty, and death presents us with the largest uncertainty we have while simultaneously containing the one certainty we cannot optimize away: You're going to die.

Somehow, some people have decided to optimize for not doing that.

Bryan Johnson is probably the most extreme public example. His entire philosophy is literally called “Don't Die.” He has organized an extraordinary amount of his life around measuring and controlling as many variables associated with aging and death as possible.

But listen to the paradox.

If you organize your life around not dying, death is already organizing your life.

It decides what you eat, when you sleep, how you exercise, what you measure, what you avoid, what you do tonight and what you don't do tonight. The thing you're trying to escape has become one of the principal organizers of the life you're trying to prolong.

And I'm no doctor or mathematician, but if the game is Don't Die, it seems worth pointing out that there is currently a 100.00% chance you lose.

(Not so) fun fact: Bryan Johnson recently announced that he has autoimmune gastritis. An autoimmune disease is, very basically, the body's immune system attacking the body it is supposed to protect.

I'm not suggesting his protocol caused it. I'm more interested in the metaphor that this presents us with: A man has built a philosophy around fighting death while something inside his body mistakenly attacks itself in an attempt to protect itself.

The protector becomes the attacker.

Which brings us back to the paradox: you can become so committed to protecting something that the protection itself begins to interfere with the thing you were trying to protect.

As I look back at my own life, I think that same paradox is hiding in all the smaller examples. The system works. And that's the problem.

It works well enough that you keep going long after you've forgotten what it was originally supposed to be for.

A faster route is useful when you're late. Sleep data can reveal real problems. Systems can remove enormous amounts of stupid work. I still optimize things.

The point is this: I no longer assume that because something can be made more efficient, making it more efficient will make it better.

Some things become worse when you remove the friction.

I know the double checkmarks on WhatsApp are useful. Read receipts are useful. Knowing that someone received your message, opened it, and ignored you at exactly 3:42 PM is objectively more information than we used to have.

But was it actually better?

There was something kind of great about getting a postcard in the mail from a friend traveling through Cambodia and thinking, Oh good. He's still alive.

You didn't know when he wrote it. You didn't know exactly where he was now. You probably hadn't heard from him in a week. The postcard had traveled halfway across the world, showed up randomly in your mailbox, and by the time you received it he might already have been somewhere completely different.

There was uncertainty built into the whole thing, and it was not something we needed to solve.

Friction isn't always a defect. Uncertainty isn't either. Maybe the parts of life we're so desperate to engineer out of the system are sometimes the parts that make the system worth being inside.

The perfectly optimized life still ends.

So I'm increasingly less interested in how efficiently I can get through mine.

I've been moving some of my own work in the opposite direction lately. Slower. More physical. A little less predictable. I'm becoming interested in what comes back when you stop trying to remove all the friction and uncertainty.

We could probably use a little more of both.

I'm playing my part.

I'll show you what I mean in October.

Stay smart,

Benoit.

P.S. If you haven't joined the October waitlist yet, you can put your name here. I could give you more information about what I'm doing, but that would undermine the last 1,800 words. You'll know soon enough.