Art and Science

If Leonardo da Vinci was never born, the Mona Lisa would not exist. Not a different version of it. Not something close. It simply would not be. If Newton was never born, someone else would have discovered gravity. Maybe ten years later, maybe fifty, but the apple still falls. The law is sitting there waiting to be found regardless of who finds it.

Science is discovery. Art is creation. Discovery is inevitable because the thing already exists and you're just the first person to describe it. Creation is not inevitable because the thing doesn't exist until you make it and there's no guarantee anyone else would make the same thing.

Some companies are discoveries. The market conditions are right, the technology is ready, the need is obvious. If this team doesn't build it, another team will build something nearly identical within a year. Cloud storage was going to happen. Search engines were going to happen. Online payments were going to happen. The specific founders shaped the execution but the category was inevitable.

Other companies are creations. Nobody was asking for a site where you post 140 characters to strangers. That's not a discovery. That's an act of creation. If Jack Dorsey doesn't have that specific idea at that specific moment, it doesn't happen. Maybe something vaguely similar shows up eventually. Maybe not.

Some of building a startup is science. Market research, competitive analysis, sizing TAMs, optimizing funnels, A/B testing pricing pages. These are all discoverable. Learnable. Someone can teach you the frameworks and you can apply them and get predictable results. This is the part that AI is about to eat entirely. It can do market research, write code, analyze competitors, build MVPs in a weekend. The science of startups is getting automated.

Some of it is art. The decision to build something nobody asked for. The conviction that people will want a thing that doesn't exist yet and that they can't articulate wanting. The taste to know what it should feel like before anyone's felt it. You can't A/B test your way to a creation. You can't framework your way to something that doesn't fit any framework.

Most founders I know think they're building creations when they're building discoveries. They think their idea is unique when really they're the third team this year to pitch the same concept to the same VCs. The market was going to produce this company with or without them. And that's fine. Discoveries can be huge. Google was a discovery. But you need to know that's what you're doing, because the playbook is completely different.

The rarer and more expensive mistake is the opposite. Treating a creation like a discovery. Moving fast when you should be moving carefully. Optimizing for speed when you should be developing taste. Racing to ship when you should be sitting with the question of whether this thing you're making has any reason to exist beyond your own conviction that it should.

A discovery company is a race. Multiple teams are converging on the same idea simultaneously. Speed and execution are everything. A creation company is not a race because there's nobody to race against. The risk is that you're wrong about whether anyone wants this at all. Creations need taste and patience. Discoveries need speed and execution. Using the wrong approach kills both.