A 2012 text where I'd already written today's doctrine

I found a text I wrote in 2012, for a roundtable at the Ambrosetti Forum, inside a project I'd called Lead the Future. I reread it today, fourteen years later, wearing a holding of AI agents and a pile of post-mortems. And my blood ran cold, in the good way: it was all already there.

I had no idea, in 2012, what an AI agent or a kill switch was. But the doctrine I run Gargency with today bears an embarrassing resemblance to that of a thirty-year-old me writing about startups for an audience of important people.

What I wrote (and never stopped thinking)

The text starts from a textbook motto — "start small, think big" — and then turns straight to the theme that still obsesses me: failure. I'll quote myself, because it's quite something:

"In a very innovative and high competitive environment, learning-by-doing is fundamental but mostly learning from your own mistake is necessary. To do so, you have to reach your breaking-point, you have to fail in order to understand how it works!"

And shortly after, the line I could paste as-is into my company's Constitution today:

"The whole concept of failure obtains a new positive meaning driven by a precise objective: acquiring information and experiences as much as you can. An iterative and incremental development based on high-frequent feedbacks."

Iterative and incremental development based on high-frequency feedback. In 2012. Fourteen years before I had agents to practice it with every day.

The "re-startup process"

The part that surprised me most is the end of the text, where I talked about what happens when a startup fails. I wrote that a startup's potential value should be treated as a common good, bankruptcy included: not just paying creditors, but protecting the skills and knowledge accumulated, so they become the platform for the next startup. I called it the re-startup process.

Today it has a more modest name: post-mortem. A project dies, and before burying it I extract its skeleton — the lessons, the templates — which becomes the scaffolding of the next one. The party app died the day it was born, and the next day the templates that came out of its autopsy were already the first version of a framework for the companies that came next. The 2012 re-startup process, in a 2026 sauce.

Why I'm writing this

Because real personal branding isn't inventing a story that's coherent in hindsight: it's realizing, looking back, that the coherence was already there, in different forms. The five-a-side with standings, the radio for thirty people, the tournament with play-by-play, the Ambrosetti text: they're all iterations of the same me. Someone who believes you learn by failing fast, that small things should be treated as big, and that what matters about a failure is what you take away from it.

In 2012 I wanted to lead the future. I hadn't understood that the future isn't led: it's iterated. One step at a time, one failure at a time.

My thirty-year-old self already knew. He was just waiting for the tools.