Ben Horowitz

Cursor built 14 different models just to understand how a real developer works. That's not a 'thin wrapper on GPT.' The same phrase was used in the '80s — 'thin wrapper around an RDBMS' — for what became Salesforce.

You don't make people great. You find people that make you great, that make the company great, that you learn from — not the other way around.

Hesitation is the most destructive thing a CEO can do. Both options are bad. That's the point. The only value you ever add as a leader is making a decision most people don't like. If everyone agrees, you added nothing. You need to run toward the pain.

We went public with $2M in trailing 12-month revenue at 18 months old. The Wall Street Journal called it insane. BusinessWeek called it "The IPO from Hell." The alternative was going bankrupt. Look at how many CEOs hesitated in that moment. Most of them are gone.

"You don't make people great. You find people who make you great, who you learn from — not the other way around." The job isn't to develop your CFO into a CFO. The job is to find the world's best CFO and get leverage from them.

The myth: GPT wrapper companies have no moat. The reality: Cursor built 14 different models to deeply understand how expert developers work. Those models compound. That data compounds. "Thin wrapper on a foundation model" is how people talk when they're not looking closely.

We invest in people's strengths, not their lack of weaknesses. Adam Neumann built WeWork into the most recognized commercial real estate brand in the world. Nobody else did that. We fund what people can do, not what they screwed up. Judge players on what they can do.