As a theoretical physics grad student, I have deep knowledge of math and algorithms and things of this nature (but mostly as related to physics), and I've worke

As a theoretical physics grad student, I have deep knowledge of math and algorithms and things of this nature (but mostly as related to physics), and I've worked in the world of corporate engineering and programming, and from what I've seen -- having a lot of deep knowledge is not really the norm and it's not really the expectation. Usually, if something difficult or complicated is required, you paper over it with Mathematica or Maple or you use one of the algorithm libraries released by e.g. LLNL or you use ANSYS, things like this that do most of the heavy lifting automatically. Even in experimental physics -- people who have doctorates and loads of experience and such -- you don't typically run across anyone writing that sort of thing from scratch because the majority of problems fall into some categories which can be attacked by a lot of the fancy tools out there these days, and so training to do this stuff doesn't require the intuition and abstract knowledge that might go into understanding the algorithms behind the curtain. To some degree, this drove me away from having an ordinary career that actually pays a living wage and towards academia. The other thing is that you can learn to use the libraries in two weeks or less and it's pretty easy to declare "I have some experience with *" if you've tried it out on your own and can do something with it. And as much as I would like to see a world where people take the time to learn theory, people are lazy, and that sort of thing isn't always viable (See also Sikha Dalmia's Gandhi Rule: http://www.thedaily.com/page/2012/04/19/041912-opinions-column-buffett-dalmia-1-3/ http://www.thedaily.com/page/2012/04/19/041912-opinions-colu... - if you don't learn it, don't expect anyone else to). The consequence is that the computer is often smarter than the people using it. Somehow, I don't really see this resulting in job losses, though, unlike many people claim. The variability and the amount of problems to solve usually mean that there is someone doing some design work, even if it is not on a very serious technical level. It also opens up new possibilities for startups, and I'm sure Stephen Wolfram would talk your head off about it... As regards "cowboy nonsense": you don't always have to attach negative behaviors to ordinary things. Getting exercise and having a social life can be good for everyone, not just douchebags.