A veteran software engineer reflects on how artificial intelligence has fundamentally altered the development process. Traditional workflows required extensive upfront planning, research, and implementation before peer review. Now developers write prompts and supervise AI-generated code, shifting from creators to editors. The author compares this transition to a historical novelist becoming an editor of student-written pages.

The core issue isn't AI's coding ability—it's the loss of deep engagement. Senior developers possess institutional knowledge about legal requirements, system performance, and future feature conflicts that AI lacks. This makes human oversight essential, but the work feels less rewarding. When developers become editors rather than creators, they lose the immersive flow state that makes programming satisfying.

Drawing on Mihaly Csikszentmihalyi's concept of flow, the author notes that both novelists and programmers enter optimal experience states where hours pass unnoticed. However, editing AI output doesn't provide this engagement. The work becomes mechanical rather than creative, reducing job satisfaction.

After months of reviewing AI-generated code, the author reports noticeable skill degradation. When bugs arise or features need implementation, the temptation to let AI handle everything grows. This dependency makes developers reluctant to invest mental energy in problems they once enjoyed solving, potentially weakening their expertise over time.