If I'm building an algorithm for packing bytes into an array, inverting matrices, or some similarly technological issue, why would insights into police profiling or sexual harassment be useful?
Because most software that actually gets used on a day-to-day basis doesn't purely work on theoretical constructs, for one: even if your application just needs really basic things like names [1] or addresses [2], it will implicitly make at least some assumptions about users, which a diverse team is much better equipped to check and correct.
Yeah, I wasn't sure how to address this one. In addition to most software not existing inside a pure abstract context, things like using 3 random points to determine the quick sort pivot just sat there undiscovered for a few decades. Since hypothesis-generation is the big undefined area, it would seem that diverse backgrounds would necessarily help. Plus we're implicitly assuming that diverse backgrounds lead to sub-par work in the pure theory domain, which is kinda shitty on its own.
And I think the parallels between a solid understanding of police "kettling" as they would apply to distributed systems and/or queueing are fairly clear. But I wouldn't expect someone who thinks inverting a matrix is tricky to appreciate either.
Because most software that actually gets used on a day-to-day basis doesn't purely work on theoretical constructs, for one: even if your application just needs really basic things like names [1] or addresses [2], it will implicitly make at least some assumptions about users, which a diverse team is much better equipped to check and correct.
[1]: http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-b...
[2]: https://www.mjt.me.uk/posts/falsehoods-programmers-believe-a...