Yes, fully agree. They make absolutely no sense at all.
Considering that predicting the next word from context is the one thing LLMs are really good at, I just don't understand how none of these developments have found their way into predictive keyboards.
I've actually switched to ThumbKey and while I don't think I'm faster with it (yet), it's at least so much less frustrating and that's worth a lot to me.
Because if the electricity produced on these vessels was ten times the normal price, it would still be peanuts in the grand scheme of things.