The End of Programming as We Know It

The article argues that each major wave in computing raised the level of abstraction, hid more low-level work inside platforms and tools, and still led to more programmers, not fewer. It treats AI as the next step in that pattern. Programming is moving toward arranging ideas, workflows, and business processes, while humans stay responsible for context, judgment, and the parts that models still cannot handle well.

Reading notes

#

• Programming has repeatedly shifted from physical circuits to binary instructions, then to assembly, compiled languages, interpreted languages, web tools, and cloud APIs.
• Each shift made older skills less essential, but it also expanded access and increased the number of people building software.
• The web, mobile, big data, and cloud services created more complexity, which kept the need for programmers high.
• Large software systems already rely on automated programs at scale, and human programmers act more like managers of digital workers.
• AI now lets nonprogrammers produce prototypes from natural language prompts, but the article says this does not end programming.
• AI expands what ordinary people can do and also creates new services, new demand, and new kinds of software.
• Programmers who adopt AI tools and learn new skills will be in demand, while those who cling to older tools and habits may fall behind.
• The article uses James Bessen’s idea of learning by doing to argue that broad productivity gains depend on a stable, trained workforce.
• AI is framed as changing what programming means, from writing code to arranging ideas into executable systems.
• New roles and practices are emerging, including agent engineers and new infrastructure for coordination, accountability, and agent-to-agent interaction.
• The article says business deployment is hard because the last mile depends on understanding workflows, edge cases, and how AI fits into real processes.
• The likely outcome is more software to invent, more AI-assisted programmers, and a broader reinvention of programming rather than its disappearance.