Stay small for
reasons that aren’t rhetoric.

There is a default way to grow a software company. You raise. You hire. You add layers. You add managers of managers. You graduate from "everyone in the same room" to "the engineers don’t see the people they ship to" inside a year.

It works, in the sense that it doesn’t kill the company. It also kills the thing the company was trying to become. The product gets duller, the writing on the website gets blander, the team meetings get longer. None of this is on purpose.

Ve is a personal-software company. The product is a model that takes on the shape of one human at a time. The internal counterpart of that — the company that builds the model — has to take on a similar shape. A bureaucracy can’t make personal software because the shape leaks.

So the default playbook isn’t just suboptimal for us. It’s actively corrosive to the product.

(a) Layers. We are trying to keep the company flat for as long as we can — not on principle but because the product gets sharper when the person writing the prompt has met the person debugging the SQL.

(b) Headcount as a proxy. The default measure of a healthy company is "we hired N people this quarter." We’re trying to make the proxy "we shipped N changes the user could feel."

(c) Bland writing. Every product surface, every email, every recruiting page should read like one specific person wrote it. Most companies grow into a bland corporate voice by year three. We’d rather write less if writing more means writing worse.

We don’t fully know yet. A few candidates: a company that publishes its own internal disagreements; a company that hires specifically for editing skill (the rarest skill in software); a company that eats its own dogfood at a level of intensity that feels uncomfortable. None of these are policy. They’re instincts we’re trying to keep.

The reason we publish the sketch is so the team can hold us to it.