The author suggested that the future of software development might be “the complete opposite” now that tools could improve discrete, well-defined pieces of software quickly and cheaply. They contrasted this with spending days of human effort and tens of thousands of dollars on AI computing, but did not complete the comparison.