Start ups as info-action engines
I will soon begin a startup incubation programme. The aim is to create a company worth hundreds of millions and give most of the earnings to effective charities.
This is the first time I’ve devoted significant time to making a start up a reality.
How do I envision operating day to day?
I want to create a system or ‘engine’ which systematically surfaces, tests, develops and deploys ideas and beliefs about products, idea spaces, founder-market-fit and product-market-fit.
This engine would mostly be a collection of processes my cofounder and I follow. This need not be particularly quant or IT heavy. Though AI could feature.
The ideal is a perfect information processing and action engine which makes the best use of its components, every day. My cofounder and I will be the main components, plus AI and employees as we go along the way.
An example of such processes? Morning check-in to decide what the most useful thing to do that day is for each of us. This includes discussing:
- Our beliefs about the state of things and confidence in those beliefs
- Whether we should be investing time in reducing uncertainty about certain ideas, implementing something, finding potential customers for feedback or doing something else
As AI gets more capable it can do a higher proportion of the work, keeping the process largely the same.
The best-case is this evolves into a template for a high performing and largely AI-staffed company.