Max Kanat-Alexander, an AI engineer at Capital One, argues that code quality only matters for durable systems, not disposable software. Teams building critical software must lower the cost of engineering rigor through automation and strict standardization.