"The realms of programmers and everyday people, separated for decades by an ocean of arcane know-how, are drifting closer together." That line, from a March 12 New York Times Magazine piece titled "Coding After Coders," kicked off the latest round of "programming is dead" discourse. Computer science professor Curry Guinn thinks the NYT got the story exactly backwards.
Guinn's response, published on his university blog, makes a simple argument: when AI handles the syntax, the bottleneck shifts to understanding systems. That makes computer science education more valuable, not less.
The Interdisciplinary Programmer Thesis
The NYT article quoted economist Erik Brynjolfsson saying people who generate code with ChatGPT or Claude might not call themselves software engineers, but "they're creating code." Guinn doesn't dispute this. His point is that creating code and building reliable software are different things.
He lists the CS fundamentals that no amount of AI prompting replaces: algorithms and data structures, memory management and networking, database design, object-oriented architecture, software engineering methodologies. These are the skills that tell you why the AI-generated code is wrong, not just that it doesn't work.
A Forbes analysis by education researcher Aviva Legatt backs this up from the hiring side. Employers aren't looking for people who can write Python, Legatt wrote. They need computer scientists who can "bridge the gap between technical capability and real-world application." Her recommendation: double-major in CS and something else.
Where the Real Value Lands
Guinn's most interesting claim is about who the valuable programmers of the next decade will be. Not pure software engineers. Instead: biologists building simulation engines, philosophers working on AI governance, economists designing financial systems, teachers building educational technology.
This tracks with what I've seen in the AI coding tool space. Tools like GitHub Copilot and Cursor are genuinely good at generating boilerplate and translating intent into code. They're much weaker at system design, debugging complex state issues, or knowing when the generated code will fall apart at scale.
The NYT piece wasn't wrong that the gap between programmers and non-programmers is closing. But closing a gap is not the same as eliminating one side. Someone still needs to understand what the AI built, maintain it when requirements change, and catch the subtle bugs that only show up in production.
Guinn's framing is the more useful one for anyone planning a career or education path right now: learn the fundamentals, pair them with domain expertise, and treat AI coding tools as force multipliers rather than replacements.