This manifesto argues that software has lost its value as a moat: anyone can now build, copy or replace it, as the author, who does not code, shows by having built most of the software he uses himself. It describes the current wave of generic AI applications as a gold rush in which only the sellers of the tools profit, and proposes software in which an agent works inside the application, on the same data and actions as the interface, learning the rules its user gives it. It argues that what cannot be copied is a company's well-structured data, the complex integrations built on it and the relationships of trust that give access to it, and closes on the industries where that work is worth doing. Written and published in October 2026.
No takes yet. Share an insight, caveat, or question.
Isaac Correa (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: