every spellcheck algorithm when you type a non-english name: uh oh! that's a typo! let me go ahead and fix that for you
does anyone else remember when google's facial recognition algorithm tagged a photo of black people as gorillas? racism is so often built into the technology we use every day and it's absolutely disgusting. and it doesn't have to be this way! it would be so easily avoidable! but the tech industry doesn't care. the tools built by the tech industry reflect all of the biases of the people who develop them, and they are most often built by people who don't consider non-white people with non-english names important enough to spend any additional development time to account for them
and y'know on top of that, these decisions reflect US cultural hegemony! just earlier, I tried implementing spellchecking in my program, and no matter what I tried, I could not get it to stop flagging the vast majority of non-english names as "errors". here's the readme for the library I tried to use:
it makes such a big deal about how inclusive they want this list to be. so, where did they get the majority of these names?
US census data
the tech industry by and large does not consider anyone living outside of the united states to be people. I even checked, and the difference between the final list and the file us-census.txt is only a few hundred names. and sure, this is just one spelling library, but it's one with over a million weekly downloads on npm and over 27,000 dependents on github. I guarantee that several of the websites and programs you use every day depend on this library for spellchecking, and whoever wrote this library decided that a list of names of people living in the united states is "good enough". it's maddening!
