I often see a misapplication of the first principles approach.
Individuals come in, question misplaced assumptions, and thanks to advancements in automated code production, tend to miss the core idea that a product is a loss function.
In any complex human (or machine) endeavor, there are multiple competing objectives that cannot all be maximized simultaneously.
The great part about technology businesses is that the cost of experimentation keeps getting lower.
Continuously questioning the assumptions and getting a handful of them right, divides brilliant products from the ones that end up getting abandoned.
before we rewrite the backend in rust can we at least try adding an index