Build something that matters.
Then put everything you have into it.

A strong belief about a problem worth years of your life.

Great founders hold a belief about a problem that does not change with each new technology or trend. The belief usually comes before the company. Bezos bet everything on the internet changing commerce, and on radical customer obsession, before Amazon existed. Musk believed humanity needed sustainable energy and a way off the planet years before Tesla or SpaceX. Both saw where the world was going before the market did.

So ask yourself: what do you know that others don’t? What do you believe that most people think is crazy, too early or too boring? Adding a small twist to an obvious idea (“customer-support AI agents!”) builds nothing that lasts. The companies that matter almost never look obvious at the start. Flexport bet that software and operations could change freight forwarding, the least glamorous part of global trade. Square bet that merchants the banks ignored, like the food truck and the barber, were a real market. At the start, almost nobody agreed with either. Both went on to change or create their category. That is the pattern: a bet most people disagree with, pushed hard enough to define an era.

A big vision does not mean a big product, a big launch or product-market fit on day one. It means you see something bigger than where you start, and you work towards it. The first product is almost always small and narrow on purpose, because that is how you start moving. Stripe started with seven lines of code to take a payment. The belief behind it was huge, and most people disagreed: that the internet would become the economy, that its payment systems were built for a world before the internet, and that developers, not banks, would rebuild them. The small product was the start. The belief was the destination. If you are still at the idea stage, that bigger belief is what you are looking for: something worth years of your life.

You do not need the full vision when you come to us. Some teams are too early for that. But you do need to be moving towards it. Twitter came out of Odeo, a podcasting company; the founders found the real product by building. Your belief can grow through real work on the problem. It cannot be borrowed or safe. Be flexible about the solution. Stay firm on the bet: where this goes, and why you are the one to do it.

The biggest trap today is letting AI form that belief for you. Use it to argue with you, test your ideas and challenge your assumptions. But original thinking needs an original source. We all use the same models now, trained on the same average of everything people have written. Ask one what to build, how to sell or what to believe, and it gives you the average answer. If your view of the world comes from there, everyone else has it too. Then it is worth nothing.

An LLM can sharpen your conviction. It cannot build it for you.

Distribution, and how fast you learn.

What matters is not how busy you are, but how fast you learn. How many real hypotheses can you test in a week? How fast do you change course when customers react? That speed adds up over time. A competitor can appear and overtake you overnight, so you can never slow down. Speed always mattered in startups. But racing a thousand competitors is very different from racing ten.

Your belief sets the direction. Speed only helps if it points that way. Founders close to the problem do not just run more experiments, they run better ones, because they know which ones are worth testing.

This is where AI helps. It makes every experiment cheaper and faster, so you can run more each week. What it cannot do is tell you which experiments to run when the answer is not yet in the data. That is still your judgment.

Everything else matters far less.