Bill Gates famously stated that people overestimate what they can do in one year and underestimate what they can do in ten years.
April 22, 2016. I’m in Barcelona for the day, and I write to a 26-year-old founder I’ve never met. Antoine Martin, the founder of AMO, the same Antoine who has written some of the most beautiful chapters of my own story, had flagged him to me: a cold email he’d received, so well crafted he thought I had to see it. Once again, Antoine was setting something in motion.
The founder shows up the same afternoon, calm, sharp, talking about the future of code like he had already seen it. What I only learn later: he had hopped on a train from Madrid that very morning, just to meet me.
His name is Eiso Kant. His company, Sourced. The tagline: AI ON CODE.
Yes, you read that right… In 2016!
We led the round alongside our friends at Frst, because Eiso wanted to bring AI to code years before the world was ready, and we wanted to be there when it happened. It went well enough that GitHub came knocking with an acquisition offer. We said no.
But that process is how Eiso met Jason, while Margarida was already working at his side at Sourced. Three founders in the making, and nobody knew it yet. Not even them.
Sourced didn’t survive being early, and we had to liquidate. Most startup stories end there but not this one. Eiso came back with Athenian, with a line I still love: we can’t automate code yet, so let’s empower the people writing it in the meantime. We invested again.
The truth is, by then I wasn’t betting on an idea anymore. I was betting on Eiso. He has something I’ve only ever found in the world’s very best founders: an insane level of responsiveness and efficiency. Fifteen purposeful meetings a day, five, six, sometimes seven days a week (I know, I know), without ever losing energy or focus. When a founder operates like that, failure stops looking like an ending and starts looking like a chapter. That’s why we kept backing him, again and again. That, and the rarest gift of all: he surrounds himself extraordinarily well. Margarida and Jason are living proof.
Then, in 2023, the world caught up with 2016. Athenian folded into poolside, and Eiso rolled our stake with it. He owed us nothing. He did it anyway. I called it a story of loyalty on my Substack two years ago. It still is.
Timing and Attitude are half the story. The other half is discernment.
Even if the timing is now right, AI today is cluttered, busy and crowded, with too many people chasing the same models, the same benchmarks and the same narrative. In that noise, doing things differently gets you questioned.
poolside took a different path: more discreet, but very forward-looking, one more time, while the rest of the frontier AI world moved very, very, very fast. Some called it delusion. What if it was clarity hidden in plain sight?
Ten years of the same conviction, carried through a refused offer, a liquidation and two rebuilds, alongside Jason and Margarida, without ever trading clarity for consensus. Last week, a decade and a train ride later, the deal with Nvidia made the news. You’ve read the numbers.
Not an ending. A new beginning.
Now poolside, the operating company, carries on with its mission, and the infrastructure company they’ve launched is just getting started. I can’t wait to see what this team builds over the next decade.
I like to joke that Eiso was never early. It’s the rest of us who are late :)
—
https://2lr.substack.com/p/poolside-and-a-story-of-loyalty
Congrats Jean