The structural predicament of the software engineer whose fifteen-year investment in implementation skills—rational at every step—has become a path-dependent trap when AI commoditizes execution and relocates value to judgment the old paradigm treated as secondary.
Arthur's path dependence framework illuminates the specific crisis facing experienced developers in the AI transition. The senior engineer chose a technology stack rationally when she began; each subsequent year of investment deepened specialization, expanded professional networks, raised opportunity costs of switching. The accumulation was rational throughout. And the accumulated capital—technical expertise, professional relationships, psychological identity—does not automatically translate into the judgment-oriented work the new paradigm rewards. The ability to write elegant code and the ability to judge whether code is elegant are different cognitive operations drawing on overlapping but distinct capacities. A developer who spent fifteen years writing code has developed judgment, but that judgment was always secondary to production—embedded in creation rather than exercised as independent faculty. The AI transition demands the secondary skill become primary. This inversion is precisely the kind of disruption path dependence makes most painful: the path-dependent investment does not transfer, and the deeper the investment, the more painful the recognition.