[jdietrich]

I think you define "strenuous" quite weakly. The overwhelming majority of jobs don't require very much cognitive effort, even in software. Most developers hardly ever actually do anything legitimately hard in their day-to-day work.

Proving theorems is a completely different class of work to chipping away at your failed unit tests or knocking together a CRUD app in your favourite language. Maybe a tenth of one percent of software developers routinely do work that is intellectually on a par with mathematical research; The rest are for the most part skilled tradesmen, doing work they understand relatively well.

It is universally accepted that all but the most prodigious musicians do not benefit from more than 4 or 5 hours of practice a day. They can often easily do twelve or sixteen hours a day, but the extra time is simply wasted. Once your reserves of concentration are spent, you're just going through the motions without learning anything. Most conservatories go to great lengths to persuade their students to practice less, because young musicians are often convinced that they can attain mastery through sheer force of effort.

I can sit and transcribe or arrange parts all day long. I can play from sheet music until my hands give out, all the while daydreaming about what I'm having for tea or what chores need doing. I can't usefully improvise or compose for more than about two hours at a time, or more than four hours in a day. I can feel the point at which I start playing familiar riffs rather than truly improvising; When I've run out of ideas and I'm just writing pastiche. There are composers who claim to do regular eight-hour days, but when you look deeper they invariably spend most of that day arranging or transcribing or recording into the computer, stuff that's essentially just admin.

[SilasX]

>Most developers hardly ever actually do anything legitimately hard in their day-to-day work.

I might agree with that in an absolute sense. But if we're going to talk about percentiles, and relative to the human population, very few people (percentage-wise) can even do the "easy, boring" part of programming you're referring to.

[Detrus]

It is hard to learn. But once you learn it, not that hard to do CRUD day to day. When you have to learn some new framework/library it gets harder again.

[robertk]

Well, as a mathematician, I can say the same is true of mathematical research. Climbing Mount Bourbaki is a difficult endeavour, but from the top the pastures are relatively peaceful to explore. Most papers are not groundbreaking theoretical sledgehammers, but minor updates on footnotes of a vaster theory.

[SilasX]

This matches my experience about technical, "it's complicated" knowledge in general: the hard part is catching up to what everyone knows, but once there, contributing and improving are surprisingly easy and obvious.

Just recently I was given a task of debugging a major problem with our website that was causing frequent emergencies -- and apparently, was mystifying the developers.

My lead realized I'd need to understand the website's infrastructure first, so he took about 45 minutes explaining how it all fit together and where the problem occurred. But once I got to that point, my reaction was, "Wait -- wouldn't the problem go away if you just ... didn't do $STEP at that point?"

Turned out to be the entire solution to the problem.

[SilasX]

Hey, I'm not the one you need to convince. You should be talking to VCs about your great idea to save money on programming by hiring burger flippers and turning them into mediocre programmers.

[qu4z-2]

Perhaps you missed the part where he said it was hard to learn.