This is a really interesting example of half of a dynamic that constantly plays out in industry. One half is that first-order reasoning seldom leads to the "ri

This is a really interesting example of half of a dynamic that constantly plays out in industry. One half is that first-order reasoning seldom leads to the "right" answer, and that optimization opportunities don't tend to be the obvious ones. Context matters, and optimizing for one thing almost always comes at the cost of other things. The "I can find the right answer without data because I'm smart and I have calculus/bayes theorem/whatever" folks are seldom right. The other half is that there's a lot of waste in industry because of missed opportunities for optimization, and companies that are good at going back to first principles and looking for those opportunities tend to be more successful in the long term. "We do it because of X" is often wrong, both because the initial reasons weren't always clearly analyzed (and often not quantified at all), and because the assumptions, underlying costs, etc have changed over time. We see this a lot in tech because the underlying costs change so much, but its true everywhere. These things are in tension, because first-order reasoning or "entitlement reasoning" is both often wrong, and often a great way to discover opportunities for large optimizations and new ways of thinking about problems. There's great value in rising above the context and complexity and reasoning things through using broader principles and simplified models. I suspect this dynamic will be with us forever. Falling too far in either direction leads to failure. If you ignore context, you'll be wrong on the specifics. If you get too lost in context, you'll miss big opportunities for optimization.