> We are coming to a point where software is developing so fast and the abstractions getting better that soon we will have more software written by a smaller nu

> We are coming to a point where software is developing so fast and the abstractions getting better that soon we will have more software written by a smaller number of people. This seems unlikely to me, or at least unsupported by current observations. Software tooling and abstractions have improved tremendously since the early days of programming. Going from machine code on punch cards to modern high-level languages with assorted libraries and frameworks and debuggers represents at least a few orders of magnitude productivity increase per programmer. But has that incredible efficiency gain resulted in fewer programmer jobs? Nope, just the opposite: the better the tooling is, the more useful each programmer is, so the more it makes sense to hire. This makes sense when you consider how open-ended programming is as a field: it lets you manipulate data, and it turns out manipulating data has a seemingly endless number of uses. It's not a field like, say, plumbing, where you need a set amount of plumbers, where going beyond that number doesn't really accomplish anything.