Skip to content

A Million Little Nooks and Crannies

2 min read
AIAgentic CodingSoftware DevelopmentEnterprise SoftwareVenture Capital

What gets built when software no longer has to hunt big fat whales?

Before you could solve a problem, you had to prove that it was attached to big money. Not just some money. Enough money to support a swath of engineers, designers, product managers, salespeople, managers, and everyone else required to turn an idea into reliable software and keep it alive.

The bill came due first. The software came later.

And before either one, a story had to be told to investors—a full-time job unto itself. Things like:

  • The market is enormous.
  • The buyers have enormous budgets.
  • The product could become a platform.
  • The largest companies in the world would eventually need it, and become addicted to it.

This was not just marketing laid on top of the software company. It was part of the machinery required to finance the software company in the first place.

This did not merely determine which software companies succeeded. It determined which kinds of software were attempted at all.

Agents have made this approach to building software prehistoric almost overnight.

Agents reverse this sequence. They can affordably participate across implementation, testing, design, integration, deployment, and maintenance, allowing one person to attempt work that once required a team. The initial bet becomes dramatically smaller.

And when a large pool of upfront capital is no longer required, neither is the story once needed to obtain it. You do not have to prove that a magical market of Fortune 500 buyers exists before solving the problem at hand. You can build something, give it to real users, and learn who needs it.

Enterprise may still be the answer to who needs this?, but it no longer has to be the premise.

The biggest shift may be who gets to decide what deserves software. The person closest to a problem no longer has to convince investors that it is enormous. They can simply try to solve it.

We have spent decades learning not to see frustrations too small to support companies as software problems. Agents will not only help us solve more of them. They will teach us to notice them.

There are a million little nooks and crannies waiting to be filled.