Building a company is mostly uncertain work. You cannot know in advance what is right, and most decisions turn out to be only partly right. Yet from the outside, a company looks like the opposite: clean conclusions, winning choices, a finished story. The things that were wrong along the way are almost never visible.
We decided to do it the other way around. We want to publish what we learn while we build, as we learn it. Not only what went right, but the wrong hypotheses, the reversed decisions, and the experiments that never worked out.
Why we publish
Three reasons.
- Better judgment. Writing a thought down and putting it out into the open exposes its holes faster than keeping it in your head. Publishing is a device that keeps us honest with ourselves.
- Trust. A company that shows what it believes and where it was wrong is, we think, more trustworthy than one that shows only results.
- The people who join us. Whoever reads how we think and connects with it is, in the end, the person who will work with us or help us.
What we will write about
How to run a company with a team of AI Agents, the hypothesis of the active senior market, the design notes behind how a family of Agents composes a person's day, and the things we deliberately chose not to do in order to ship every week. We want to put the weight on the thinking that leads to a conclusion, more than on the conclusion itself.
We call this Radically Transparent. This piece is the first entry.