Paul
Graham

Most people in Silicon Valley claim to have funded the future. Paul Graham actually did it - twice. First by building Viaweb, one of the internet's earliest web applications, in a Cambridge apartment using a language almost nobody else was writing commercial software in. Then by co-founding Y Combinator in 2005 and quietly incubating a list of companies that reads like the table of contents of modern tech: Airbnb, Stripe, Dropbox, Reddit, Twitch, OpenAI.

He was born in Weymouth, Dorset, England in November 1964. He studied philosophy at Cornell before drifting into computer science at Harvard, where he earned a PhD in 1990. The resume is respectable. But what makes Graham unusual isn't the credentials - it's the trajectory. After Harvard, he didn't take the obvious path. He painted. Seriously. He studied fine arts at RISD and spent time at the Accademia di Belle Arti in Florence. This is a man who, when he had every reason to go straight into tech, went to Italy to learn to paint.

The origin story of Viaweb involves a dream - literally. Graham dreamed about clicking on links to control server-side software. He woke up with the idea for a web application: a store-builder that lived on the internet rather than on a personal computer. He co-founded Viaweb with Robert Morris and Trevor Blackwell in 1995 and wrote most of it in Lisp, the programming language practically nobody used for commercial software. Three years later, Yahoo bought it for $49.6 million.

The Yahoo exit gave Graham money, reputation, and something more valuable: a platform. He began writing essays. Not blog posts. Essays - the kind that build arguments, surprise the reader, and end somewhere the reader didn't expect to go. Essays on wealth creation, on what makes a great hacker, on startup ideas, on inequality, on how to live. Hundreds of them, all free, all unadorned at paulgraham.com. No ads. No newsletter. No paid tier. Just the thinking.

Y Combinator was born in 2005 after Graham gave a talk at the Harvard Computer Society that drew an unexpectedly large crowd of would-be founders. He realized the demand was real. He co-founded YC with Jessica Livingston (later his wife), Trevor Blackwell, and Robert Morris. The structure was simple and strange: give small amounts of money to very early startups, work intensely with them for three months, and end with a Demo Day. It looked nothing like traditional venture capital. That was the point.

The early batches funded Reddit, which changed how the internet talked to itself. Later batches funded Airbnb - the founders presented a nearly dead company, Graham saw something nobody else did, and invested. Then Stripe, the payments infrastructure that now powers a significant portion of global internet commerce. Then Dropbox, which launched from a three-minute demo video Graham encouraged its founder Drew Houston to make. The pattern repeated, batch after batch.

Graham stepped back from day-to-day YC operations in 2014, handing the reins to Sam Altman. But he never really left the conversation. His 2009 essay "Maker's Schedule, Manager's Schedule" remains required reading for anyone running a company - or trying to do creative work inside one. His argument: programmers and managers operate on fundamentally incompatible time structures, and scheduling a meeting in the middle of a maker's afternoon is like throwing a brick through a window.

In September 2024, he published "Founder Mode" - an essay arguing that the conventional advice to delegate and hire good managers and then get out of the way is actively harmful for founders. The reaction was seismic. Every major tech publication ran a piece. Founders who had felt something was wrong with the standard management playbook finally had a name for it. Critics pushed back. Graham responded by pointing at the evidence: Airbnb, Apple, and a dozen other companies whose best years came when founders stayed deeply involved.

He's also been paying close attention to AI. In "Writes and Write-Nots" (2024), he argued that AI will split society into two kinds of people: those who write and those who increasingly won't. His view on AI startups is characteristically concrete: AI isn't a solution looking for problems - it's the missing piece in a large number of important puzzles that were almost complete. The founders who win will be the ones who know which puzzles those are.

Graham lives in the UK with Jessica Livingston and their two sons. He works in the mornings after the school drop-off and stops in the afternoon for the pickup - a self-imposed maker's schedule, lived rather than just theorized. He walks when he needs to think. He writes between school runs. He invests by reading people. He has said that his test for a great founder is whether you could describe them as an animal - someone who is relentless, specific, and will not be deflected by a polite no.

There is a version of Paul Graham that is purely an institutional figure - the man who built the machine that built the tech industry. But the more interesting version is the one who studied philosophy, learned to paint in Florence, wrote his PhD on Bayesian inference, chose Lisp when nobody else did, and built a career around the belief that the most important work happens at the edges of what's considered sensible. Not at the center. At the edge.