but then it kinda stops, couldn’t we layer on chords as well, you can peck “t h e” or just mash “eth” and both would result in “the”, or maybe some sort of english pinyin where only the first few chars are needed?
Predictive text? But it's also an example of an incremental improvement that someone could experiment with to see if it works. Much easier than having to prototype entirely new hardware (especially in the 80s, in an era before cheap additive prototyping).
I mean in some ways that is what modern predictive text is, right? I watch others rely on it incredibly heavily, no correcting mistakes at all just rely on the autocorrect to solve it, and it works well
Or swipe-typing, as another version of that idea. Move in the rough area of the key, and rely on prediction
