Two examples:
1. No matter how many times I type a student’s name, Shaoni, in Gmail, Gmail’s autocorrect always wants to convert that to “Shane.” Ditto for Facebook. Shaoni is a South Asian name. Shane is a name popular with those who revere American westerns, etc. (There’s a famous YA novel and movie called Shane from the 1950s, about a renegade loner-hero with a heart of gold.)
2. Gmail and FB do the same with the last name of a Latino author I’m teaching this semester, constantly “correcting” Urrea to “Urea” (which means urine!) despite repeated corrections of this “correction” made by me. (Luis Alberto Urrea, author of The House of Broken Angels, The Hummingbird’s Daughter, and other novels.)
Wouldn’t it be possible to add some simple AI where if a suggestion were rejected more than 3 times by the same user the system could “remember” that? Just a (hopeful) suggestion. (Yes, I know it’s possible to reject suggestions by clicking on the little X button. But when you’ve got lots of typing to do and email etc. to get through, having to reject repeatedly wrong prompts really is a waste of time.)
We all know that Google and FB have other quirks (to use a too-mild word) that are more damaging to our society than imperfect autocorrect algorithms. But how about using some of their millions to pay a few bright 20-something coders to fix these problems? They’re not, actually, unrelated.
Or am I mistaken and all this is an Apple autocorrect issue? (I use a MacBook Pro.)