The Future of Software Looks Like Services

When building app software was hard, there was a moat in software. The products weren’t easily replicable and that was enough. As building app software becomes easy (or even automated) the execution, biz logic, systems, etc. become the moat but that’s very hard to prospectively underwrite (without proof) even if it’s relatively easy to identify at scale. So pure application software businesses are increasingly uncertain to start or invest early in early. Eventually/soon we’ll get to a world where you just ask GPT-10 to “build me an app to do X, has these features, and looks like this screenshot” At that point, there will be no moat and infinite competitors for application software. Software companies (and their margins and outcomes) will generally compress to look services companies: easy to start, usually not very big, little or no need for outside capital at inception, etc. Because there is increasingly no moat in software, building a big high-quality business requires finding someth

When building app software was hard, there was a moat in software. The products weren’t easily replicable and that was enough. As building app software becomes easy (or even automated) the execution, biz logic, systems, etc. become the moat but that’s very hard to prospectively underwrite (without proof) even if it’s relatively easy to identify at scale. So pure application software businesses are increasingly uncertain to start or invest early in early. Eventually/soon we’ll get to a world where you just ask GPT-10 to “build me an app to do X, has these features, and looks like this screensho