The difference between academia and practice is that academia is only satisfied with a perfectly coherent theory and practice is perfectly satisfied with an inc

The difference between academia and practice is that academia is only satisfied with a perfectly coherent theory and practice is perfectly satisfied with an incoherent but working solution. Sometimes, it's more productive to use words, ideas, and objects that already exist to solve problems. Even if they're not ideally suited. To answer the "filing taxes is needlessly complicated" actual problem, the "overlearning" (by which they seem to mean "over-optimizing for risk reduction") explanation would be that the individuals in government responsible for writing the tax code have all optimized for actions which will further/protect their careers. And those actions inevitably lead to a more complicated tax code, because {adding tax code} generally makes important entities happy (exclusions! or more heavily targeting not-you!) and {removing tax code} only makes the people happy (in an abstract way, albeit often angry in a specific way). An example of government policies that could help fight this are the "1-in-2-out" style regulatory reduction mandates. Where in order to implement a new regulation, previously standing ones have to be culled, which effectively incentivizes pruning useless or less important regulation.