In academia, the goal is the way. Elegant solution is much better than the straight and complicated solution. Even even if you don’t solve the problem, failures are also appreciated.
In business, the goal is the solution. Have it complicated, have it simple, have it full of bugs- as long as you have final solution.

This is why academia is reffered to as the ivory tower- you don’t have to actually solve a problem, you just need to have something new. This is why business is sometimes so frustrating- it is okay, and even preffered, to copy something that works, but you must solve the problem. Elegant failures or clumsy solutions- take your pick.

Chris Lomont ask himself why does the code works, and how the constant was first found. He described how does the code works, but failed to explain how the constant was found. During the process, he found that 0×5f375a86 was a slighly better magic number than the original one.

While Carmack can neither be labeled Academia or Business, Chris Lomont’s paper is clearly an academic paper. That’s why I think that xkcd note refer to Chris Lomont study.

Ho, and by the way, it appears that John Carmack is not the father of this code

http://www.beyond3d.com/content/articles/8/

It look like noone ever wrote this code, while every suspect remember that code, each one says he’s not the original author. Clearly similar like all the javascript code written between 1995 and 2005, only copied from other web sites, never written from scratch.

Lots of “I think”s, it’s a while back and I’m a little vague on the subject, but it was something like this.

The hex number is from this algorithm for taking the inverse square root of a number. It uses an iterative algorithm but uses an ingenious first approximation to make the result tolerable after only one iteration.

Magic number in an inverse-square-root approximation algorithm.