“Software has always been more expensive than we can afford to build. There are entire categories of applications that don’t exist simply because the economics don’t work.” When migrations become 10x faster, security fixes 20x faster, and test coverage jumps 40 percentage points with minimal effort — suddenly a lot more software becomes economically viable.
This is the supply-side argument for why Technology transitions create more of the 'dying' thing, not less. Sinofsky’s historical pattern explains that transitions always create more of the “dying” thing — Dabit explains the mechanism: cost reduction doesn’t just make existing software cheaper, it makes previously impossible software possible. The constraint on software development shifts from writing code to knowing what to build.
Dabit also draws a sharp distinction between model labs (train foundation model, then figure out product) and agent labs (build tools that solve problems, then invest in models). Agent labs are product-first. This connects to why In agent-native architecture, features are prompts — not code matters architecturally — agent-native products compose capabilities rather than coding them, which is how you get the 10-20x cost reduction that enables software abundance. The implementation gap collapsed for individual builders, but the abundance thesis says the same thing happens at the industry level.