AI Is Making You a Worse Engineer and a Better Employee

Somewhere in the last few years, the meaning of being "good at your job" and "good at your craft" for software engineers started to diverge, and most people failed to recognize it. AI coding assistants have allowed developers to ship faster than ever before, creating whole functioning prototypes in minutes, closing tickets quicker, and appearing as stronger employees based on manager metrics.

However, the underlying skill that this productivity used to signal - system-level intuition, debugging instinct, and the ability to reason through unfamiliar problems independently - is being eroded by these tools. As these "muscles" go unused, the underlying skill quietly deteriorates. The issue is not whether AI is bad or should be abandoned, but rather that we have been measuring the wrong thing - mistaking throughput for growth.

The assumption that productivity and cognitive development are the same curve has been shattered by AI, which makes the gap between them clear. Productivity measures output, while cognitive development focuses on the growth of reasoning abilities when faced with new problems. AI has severed the link between productivity and cognitive development, allowing developers to ship more while solving less.

This shift in focus threatens to degrade problem-solving skills that were previously implied by productivity. The danger lies not in AI slowing down developers or making them less effective, but rather that measuring productivity alone does not reveal the erosion of crucial skills. It is essential to recognize that a decline in problem-solving ability may be hidden under increased productivity, and the true measure of success should go beyond just output.