AI coding assistants promised to make software development faster and easier. Tools like GitHub Copilot and Cursor autocomplete functions while agents scaffold entire features from plain English descriptions. The barrier to producing code has never been lower.

Yet software engineers report their jobs have become more complex and exhausting than ever before. A February 2026 Harvard Business Review study of 200 tech workers found that 83 percent experienced increased workload after AI adoption, with 62 percent reporting burnout. The baseline for what engineers are expected to deliver has quietly shifted upward.

Engineers now face expanded responsibilities across product thinking, architecture decisions, and code review while being told to write less actual code. The identity shift from builder to reviewer has created a professional crisis for many who entered the field to write code, not manage AI-generated systems. As one engineer described it: every day feels like judging an endless assembly line of pull requests.