Even though we often leave the brute force calculation to machines, we still need to understand the underlying theories at least well enough to program and operate computers properly. Have you ever been at the checkout and had a cashier who can’t count out correct change? If so then I think you see what I mean.
And then there are the slightly more practical applications. For instance, I work on cars. Sometimes I work on American cars that use SAE wrenches/sockets. Now, if 5/8” is a hair too small, do I go with the 19/32”? I happen to know fractions and be able to do them in my head well enough to know that that would be a waste of time; the 21/32” would be the right choice. That is just a simple example.
When I am working in a machine shop, things can get a bit more complex. Try figuring out the maximum RPM you can run a tool without burning it up due to a too-high surface speed and then figure the optimum feed rate at that RPM without causing excessive tool wear based on the RPM, the number of cutting edges, and the optimal chip size and you wind up doing a little geometry (Tool_circumfrence=pi * tool_diameter), unit conversion (to convert inches per revolution to surface speed), and simple math (to calculate feed per revolution and then combine that with RPM to come up with a linear feed rate).
That is part of the reason machinists get paid pretty well; that sort of math skill, while actually simple, is increasingly uncommon in this age where the average HS student is more concerned with finding proxy sites to log into Facebook from their school computer than actually learning.
Accountants need math, engineers need math, cashiers need math, anybody who gets paychecks or bills needs math… pretty much everybody needs at least some math and to understand at least some of the principles behind it. If nothing else, computers are not infallible!