by Vivianne » Fri Oct 11, 2019 12:35 pm
I saw that article, Lemon! The exoskeleton, right?
Okay, here’s how I -tried- to type during lectures when my dominant hand was declining.
I mirrored the QWERTY keyboard manually so that each key could represent one of two letters. Q could have been q or p, w could have been w or o, and so on.
It was a pain to interpret, so I quickly modified the system to include a Visual Basic script that would change capitalized letters exclusively. That led to much abuse of the caps lock key, and it wore on my hand.
I graduated shortly afterward and don’t really use it anymore. But AI now could make it interesting! I’m not a huge tech person, but one of my brothers is possibly going into AI and has expressed an interest in accessibility. If it ever comes up, I’ll see if he wants to refine this method. I have a starting point for him; I once did an ngram analysis of the Google books corpus after seeing Peter Norvig’s letter frequency analysis.
Anyway, maybe something based on my method could be made helpful, if basic efficiency obstacles can be overcome.