I've taken to calling it autoincorrect. The prediction model for phone text seems to only consider word frequencies and blindly replace "were" with "we're" or i

I've taken to calling it autoincorrect. The prediction model for phone text seems to only consider word frequencies and blindly replace "were" with "we're" or its with "it's" every time it is encountered without regard for grammatical context. It's incredibly frustrating to type the correct word only to have it mangled by an algorithm that is less sophisticated than MS Word 95.