bullgod66

If you wrote a new library, what would you put in it? What is the point of writing a new library that contains functions that already exist in other libraries?

I'm not saying industry doesn't conceptualize unmet needs. I'm only saying that it's intellectually uninteresting.

What is intellectually interesting about just copying and pasting code from a library? I think we can all agree that it is significantly (orders of magnitude) less interesting than solving abstract math problems.

Software development is not interesting - you don't have to come up with any novel ideas, or be creative in your problem solving, especially with the existence of large libraries that already exist that can perform any algorithm for you. For example, if you are doing machine learning in Python, you can just use a library like scikit or numpy, and you don't even have to code the algorithm. In fact, you don't even have to understand the algorithm or why it works - your boss just tells you to use k-nearest-neighbors and you select that from the library - no independent thought required. It's basically mindless - I don't understand what is intellectually interesting about it.

And in order to do interesting mathematical work in industry, you have to have a PhD and therefore you must again be placed in the extremely competitive and toxic environment of academia. And regardless, any theoretical knowledge of pure math would be unnecessary due to the existence of these libraries that basically do all the thinking for you. Even if you are smart and able to do pure math (which should give you a competitive advantage), it doesn't.

Although this article makes a good point, it doesn't offer any practical suggestions for how to avoid the issue entirely.