I've Been Running for 16 Years. I Still Haven't Arrived.
From 4:26 to 3:19 in less than a year, and an injury that will probably keep me out of marathon number 20. Why someone who does an endurance sport understands something about surviving AI that someone who only reads rarely gets.
Sixteen years ago I started running. Five months later I lined up for my first marathon with the goal of breaking four hours. Everyone told me I was crazy, that it wasn't normal to do that with so little preparation. Today you see "influencers" doing it for the camera without having run a day in their lives, but that's another story.
I ran 4:26. I finished wrecked, walking the last kilometers with cramps in places I didn't know could cramp, wondering at what point this had seemed like a good idea.
Eight months later I ran Berlin. I ran 3:19.
From 4:26 to 3:19 in just under a year. Not because talent suddenly showed up, but because between those two marathons there was something you don't see in any finish-line photo: hundreds of kilometers alone, at inconvenient hours, with nobody watching (there was no Instagram and we were happier). On many of those days I felt the effort was doing nothing. But that invisible work is what broke the stone.
I'm telling you this because I'm less than 12 weeks out from running Buenos Aires for the 5th time, what would be my marathon number 20, and most likely I won't make it. The injury that started in January is back.
And even though I write a lot about judgment and systems, today I want to talk about the other half of the equation: why someone who does an endurance sport understands something about surviving AI that someone who only reads rarely gets.
The two people I was
Before running, I was a different person. Full-stack developer and one of those guys who only goes to the gym when vacation is coming up. I did the operational work, I delivered, I floated. Sport was seasonal aesthetics and work was a list of tasks that got finished and forgotten.
At some point two things happened that I now understand were the same decision: I started running seriously, and I entered the SAP world shortly after. In one stroke I changed friendships and coworkers; a radical change, but a necessary one at that moment.
It's no coincidence they happened together. Both were the same underlying choice: stop doing things that end and get forgotten, and start doing things that build you.
Running isn't a task you cross off a list. It's a practice that changes you while you do it. And entering SAP, for me, was the same thing professionally: stop typing code that solved today's problem and get into something that demanded understanding systems, processes, and the business behind them. Stop executing and start comprehending.
The person who runs and the person who deliberately enters a complex domain are the same person: someone who decided that simply "floating" wasn't enough.
The training nobody sees is the only one that counts
Here's the bridge to AI, and it's more literal than it seems.
In the marathon, the work that transforms you is the invisible kind. Nobody applauds you at kilometer 12 on a random Tuesday. There's no photo and no medal. And yet, that kilometer is the one that makes you. The Sunday race with the crowd shouting is only where you collect what you already built in solitude.
The exact same thing happens with AI, and most people are doing it backwards.
The easy way to use AI is the operational one: ask it to solve the task, copy the result, deliver and forget. It's the equivalent of going to the gym only before vacation. You deliver, it looks good for a while, and nothing sticks. You didn't learn anything, and next time you depend on the machine just as much.
The way that transforms you is the other one: using AI as training. Not just to have it do, but to understand how it did it. Asking why it chose that approach, arguing with it, making it explain what you don't know (which is a lot, even if the machine keeps agreeing with you). Keeping the judgment and not just the output.
That's the invisible Tuesday kilometer. It doesn't look impressive and it's slower than copy-paste. But it's the only thing that, a year from now, will leave you as someone the machine doesn't replace.
Don't float just to float
There's a temptation, especially when you have years in the trenches behind you, and it's to let yourself drift. Do the work that comes, deliver, get paid, repeat. And with AI that temptation became more dangerous than ever, because now you can float faster and with a better finish. You can produce twice as much without understanding anything, and have it look good.
That's exactly the moment you become replaceable. You're digging your own grave without realizing it.
The mindset of someone who trains applies to everything. It's not about "pushing harder" David Goggins-style just because. It's about not accepting what comes down from above without filtering it. Not from the boss, not from the vendor, not from the keynote, not from the AI itself.
Always ask yourself what actually works and what is pure hype. Learn to tell the difference.
Develop critical judgment the way you develop an aerobic base: with invisible kilometers, day after day.
That's why this space exists. Above Average isn't a place to read what SAP did this week. There are a thousand newsletters for that, or AI-generated posts. This is a reminder that reading what happened isn't enough. You have to get your hands dirty, test what works, be wrong, and refuse to settle. Reading is watching the marathon on TV. Running it is something else.
What's coming
It's almost certain I won't run my marathon number 20 in 12 weeks. In your mid-forties, biology doesn't negotiate. It hurts, because I was excited about the round number this year.
But sixteen years of training doesn't collapse because one race does. The capacity is still there. Sport for me was never about a single race: it's a second job, the place where I reinvent myself and force myself to stay competitive. And that logic is the same one I apply professionally.
The marathon you trained for can be cancelled. The technology you master can become irrelevant. Your role can disappear. None of that is under your control. The only thing you control is whether you kept training when there were no guarantees.
I've been running for 16 years and I still haven't arrived. That's the point. You never arrive. You keep training and learning no matter what happens to you.
See you at the invisible kilometer.
Pablo / En Producción
P.S. If you're floating just to float, at work or in your running shoes, this is your reminder that the kilometer nobody sees is the only one that counts. Tag or forward this to someone who needs to read it today.
Newsletter
In Production
Every Tuesday: what is actually happening with AI inside large enterprises, written by someone implementing it there.