AI Wasn’t Made for Me — But It Feels Like It Was
Sometimes I joke that AI was invented for me. Of course it wasn’t — but the feeling is real. I never expected, before the 2020s, that such a technical leap would ever happen in my lifetime. Yet suddenly, here we are, and I feel… unleashed. Something has shifted, both in the world and inside me.
The mistake that cost me 30 years
At the end of the 1980s, I worked for a small CAD/CAM software editor using APL — a bizarre, elegant, powerful language. The problem at the time was cultural: what was meant to be a prototyping method became, out of greed and convenience, a production method.
Meanwhile, I spent my days writing assembly code — nobody else wanted to do it — and I loved it. That period could have changed my life.
I even had an opportunity to join SGI, Jim Clark’s company, one of the mythical pillars of Silicon Valley. The kind of place where VPs showed up in flip-flops, and engineers drove sports cars bought with stock options. A dream.
But instead of leaning into engineering, I made what turned out to be one of the biggest mistakes of my professional life: I decided I should “evolve” toward project management.
From the early 1990s onward, I stopped programming. That error lasted more than 30 years.
I paid for it with dependence: solid concepts and ideas, but never the hands to fully execute them.
I still remember, in the mid-2000s, when I patented a GSM-based system internationally. To implement it, I had to design a custom communication protocol using SMS as instruction packets. Debugging was a nightmare: developers didn’t fully grasp my vision, I lacked their technical immediacy, and we all suffered from the gap.
The Colombian genius — and the missed opportunity
Ten years later, an engineer at META introduced me to a Colombian developer — a phenomenon. A one-man army. Someone who could rebuild in two months what teams failed to deliver in four. A mentor for all, but terrible at teamwork.
He and I could have built wonders.
In the late 2010s, I saw the opportunity to build IAM and IDP systems before the passwordless wave. I tried to convince him. He didn’t see it. We missed the train — he missed the vision, and I missed the conviction to push harder.
Years later, he came back and didn’t understand why I no longer shared my AI vision with him. The window had closed. Timing matters.
AI brought me back to life
I never thought I would return to real software development. Then AI arrived.
Suddenly, all the small barriers that had kept me away from coding evaporated. My impatience — my worst enemy — was neutralized. AI gave me the ability to implement ideas at the speed I think them.
Large language models reopened a world I thought lost to me forever.
Now, the only thing I fear lacking… is time.
This is not the 2000 bubble
I lived through a bubble already — the early-2000s rush of e-commerce. And yet, today feels fundamentally different.
Some players are burning billions — the new unit of measurement in AI. But the rest of us? The real innovators? We barely need seed funding anymore. The enormous capital invested in giant AI labs is indirectly financing thousands of “modest” but ambitious founders who leap ahead thanks to these tools.
Most early-stage AI entrepreneurs today can develop powerful prototypes without asking anyone for money.
Investors, in turn, will need to base their decisions on real traction, real scalability, real users — not hype.
For once, we may be witnessing a technological wave that grows without bursting.
And me?
I feel like a beginner again — in the best possible way. Thirty years of ideas I could not implement are suddenly becoming executable.
AI didn’t free humanity — not yet — but it freed me.
And this time, I don’t intend to waste the opportunity.