Paul Ford writes in The New York Times that the initial fear of software developers being replaced by tireless robots has largely receded. He argues that while artificial intelligence can generate competent code, it also lowers the barrier to entry for poor quality work. The industry is now recognising that creating truly effective software still demands human collaboration and distinct craft skills. Ford suggests that because everyone can now write code, many should not be doing it at all. The focus has shifted from mere production capability to the necessity of human oversight in finalising projects.
This perspective highlights a shift away from volume-based development toward quality assurance and strategic thinking. It implies that the value of a developer lies in their ability to judge when to intervene rather than just typing syntax. The result is a market where low-effort coding is devalued while high-level problem solving becomes more critical.
- AI makes bad code easy to produce
- Human collaboration remains essential for quality
- Many new coders should not write production code
