When a company competes on a function, falling execution costs don’t shrink the work. They expand it: as the cost of building falls, the cost of not building rises, because every competitor is building more too.
Software companies will build more software than ever, and because AI moves validation downstream, they’ll test more ideas by shipping them. More execution, in turn, needs more coordination and more judgment about what to build, for whom, in what order, and whether it’s any good.
This only holds for advantages. For a necessity, cheaper execution just means reaching “good enough” sooner, because necessities have a point of diminishing returns. The same drop in cost grows the work in one company and shrinks it in the next, which is why whether a function is an advantage depends on the company.