We build businesses that run entirely on software, and we own all of them.
For all of history a business was a structure made of people. Financing, valuation, hiring, org design, all of it follows from that. You could not start fifty businesses because each one needed a person inside it.
That constraint is now optional. Starting a business stops being an act of assembly and becomes a production process, with a rate.
When trying costs almost nothing, what is scarce is knowing what to try, and being willing to kill what does not work.
An idea goes in. A business comes out with its own strategy, product and customers, and nobody inside it. Every job a company would hire for is done by software that keeps rewriting itself.
A short list of decisions stays human. The system decides and carries out the rest alone.
Every business runs on the same operating system. When one works out how to do something, the knowledge goes into the system instead of leaving with the business. Everything built afterwards starts with it, and the ones already running receive it too.
A tenth business is not much cheaper than the first, because the first was already cheap. It is far more capable.
The early ones were built to find out what the system could not yet do. Each exposed something missing, we built it, and the next one started with it.
Five were retired on purpose once they had taught us what they were for. That is how the cost of being wrong stays small enough to keep being wrong. The two still running have been operating since March.
No revenue yet, and none was built to earn any. The next group is the first that has to.
The aim is hundreds. Each owned outright, each run by the same machine, each one leaving that machine better for the rest. Most will not work, and that is priced in. The ones that do are owned completely.
Built Round One and ran it for a decade, sold it to Info Edge, then led generative AI there. Johns Hopkins, INSEAD. LinkedIn