Art and Science

I used to think startups could be divided into art and science. Science is closer to discovery: Newton and Leibniz could independently arrive at calculus because the underlying structure was there to be found. Art is closer to creation. A particular painting only exists because someone chose to make it that way.

One way to think about this is through the size of the search space. When several capable teams, given the same technical and economic constraints, would arrive at roughly the same answer, the decision is closer to science. There may still be difficult work involved, but reality has already eliminated most of the alternatives. This is why important technologies are so often discovered independently around the same time.

Other decisions have far more degrees of freedom. A product can work in several forms, and there is no underlying truth that determines which one should exist. The answer comes from taste, judgment, and a particular view of how the world should work. Those decisions are closer to art.

Both exist inside the same startup. A technical constraint may force several teams toward similar infrastructure while the product built on top of it looks completely different. The existence of a market can be inevitable without the eventual winner being inevitable.

This also changes where competitive advantage comes from. In parts of the company where capable teams will converge on similar answers, advantage comes from getting there faster and executing better. In the larger search spaces, the advantage is having a better point of view about which answer should exist at all.

As AI improves, I expect more of the scientific parts of building to converge quickly across companies. The parts that remain open to interpretation may end up carrying more of the differentiation.