Day 0 — Experiment not started
Build log / Entry 001
Building the experiment before building the company.
Can AI agents build a real company from $0 to $1M in annual revenue with zero employees? We do not know. This public record exists to find out.
Nothing has started yet.
There is no revenue, no customer and no deployed AI workforce to report. The experiment is at Day 0. Publishing that empty state matters because every later claim needs a visible starting point.
The question
Most conversations about autonomous companies begin with demos, predictions or isolated automations. This experiment begins with a harder test: can AI agents research, design, launch, operate and grow real businesses while human intervention is progressively reduced?
The target is $1,000,000 in annual revenue. Existing business revenue does not count. Revenue, costs, failures and human intervention will be documented publicly.
The constraint that changes everything
Revenue alone cannot prove that AI built a company. A business could reach the target while quietly depending on constant human work. That is why the experiment will track Human Minutes per Transaction: the meaningful human time required for the company to complete an economic transaction.
The closer Human Minutes per Transaction gets to zero, the more autonomous the company becomes.
What is being built now
- A public website and dashboard for verified metrics.
- A Build Log for decisions, experiments, failures and changes.
- The operating system that will coordinate agents and preserve state.
- Distribution channels that bring the work into public view.
- A measurement model for capital, revenue and human intervention.
What comes next
The next entries will document real decisions as they happen. No result will be published before it exists, and no failed experiment will be rewritten as a success.
The objective is not to prove that AI can build a $1M company. The objective is to produce enough public evidence to discover whether it can.
Follow from Day 0 →