Paul Graham didn’t set out to build the most influential startup accelerator in the world. He just hated bad software. A computer scientist with a painter’s soul, Paul made his first tech fortune by creating Viaweb in the ’90s—one of the earliest SaaS products, long before that term existed. He sold it to Yahoo in 1998, but it wasn’t the exit that changed his life—it was the essays.

Graham started writing long-form reflections on startups, programming, and human nature. They were dense, meandering, uncomfortably honest. But they resonated. “I wasn’t trying to be a guru,” he once said. “I just wrote the things I wished someone had told me at 22.”

In 2005, he and a few friends launched a small experiment: fund a handful of technical founders with just enough money to survive and enough feedback to get better. The result? Y Combinator.

At first, it was summer camp for hackers. But something clicked. By removing gatekeepers and betting on raw curiosity, YC began surfacing founders who’d never have passed a Sand Hill pitch meeting. Reddit, Stripe, Airbnb, Dropbox—all early YC companies.

What made Paul different wasn’t just his technical background. It was that he saw founders before they saw themselves. And he had a gift for language—naming the unnamed instincts of builders with terms like “ramen profitable,” “default alive,” and “make something people want.”

“I’m not a businessman,” he once shrugged. “I’m just good at noticing who’s real.”