The widespread fear that AI will devastate knowledge economy jobs ignores how previous technological waves actually played out. While reports focus on what AI can do, history shows employment outcomes depend on factors like demand elasticity, regulatory barriers, and second-order effects that are nearly impossible to predict.

Software development exemplifies the demand surge effect: productivity surged since the 1990s yet web development employment rose dramatically because appetite for software far outpaced labor savings. Similar patterns emerged in accounting, architecture, and healthcare diagnostics — AI now outperforms radiologists, but regulations and insurance requirements effectively prohibit fully automated radiology, insulating jobs.

The contrast with manufacturing is stark: saturated demand means productivity gains translated to falling employment. Retail tells an even more complex story — the internet hollowed out store jobs while boosting warehousing employment. Spreadsheets devastated bookkeeping while creating more accountant positions. Bank tellers survived ATMs but were undone by smartphones.

The question "Can AI do this task?" offers limited insight. Thus far, occupations most exposed to AI are as likely to have grown as shrunk — suggesting the jobpocalypse narrative gets the story fundamentally backward.