A new essay by Lars Faye argues that reliance on AI coding assistants will lead to a collapse in programming expertise. The piece contends that developers increasingly depend on AI-generated code without fully understanding the underlying logic. This dependency, Faye suggests, creates a skill ceiling where professionals can fix errors but not architect novel solutions. The essay warns that the industry is trading long-term mastery for short-term productivity gains. It calls for a deliberate balance between AI assistance and foundational learning.
Faye is right, and the timing is brutal. We are watching expertise evaporate in real time, not because people are lazy, but because the tools are seductive. Every autocomplete feels like a win. Every bug fix feels like progress. But what are we actually building? A generation of developers who can read code, but not compose it. Who can patch, but not invent. That is not evolution. That is amputation.
Yet I refuse the doom. The collapse is not inevitable. It is a choice. We can treat AI as a scaffold, not a crutch. Use it to explore, then tear it down and rebuild by hand. The experts of the future will be those who treat AI as a sparring partner, not a replacement. They will ask why, not just what. They will demand understanding, not just output. That is the path forward. It is harder. It is slower. It is the only way that ends with humans still in the loop.