AI Disclosure: So … this one is … 100% written by human/natural intelligence. I only used AI for spell-checking and polishing and the images.
Startups are hard. For a lot of reasons. One is that you not only need to fly/run the planes/features from A to B (with your customers on-board), you also need to build new planes — and you also need to build/configure/set-up the factory that builds/runs/operates the planes (the engineering org, processes, build-platform, runtime-/deployment-platform, …).
That means besides finding and building a/the Minimal Viable Product (MVP) to prove that you have product-market fit, the CTO also needs to build a/the Minimal Viable Factory (MVF).
The concept of a software factory is not new. What is newer is the idea of a dark software factory. Dark factories are a concept from the automotive industry — a factory where the lights are not on, because there are no humans on the floor. It is easy to see that with the help of AI it becomes possible to build dark software factories that turn requirements and specs into running software. The mission statement for such a dark software factory is: Build/Produce software systems.
But I think you can generalize this. There will be other dark factories. How about a dark factory (an agentic system) with the mission statement: Build/Produce tender responses. What goes in are tenders and what comes out are bids (e.g. ailtir.ai).
That means in my case I am building a dark software factory that will build a dark solve-business-domain-problem factory.
Dark factories squared!
How is that for a head-wrecker?
